text
stringlengths 0
383k
|
---|
ApateWeb Malicious Redirection Campaign
A recent report by Palo Alto called ApateWeb covers a malicious redirection campaign which takes unsuspecting users to websites hosting malicious content. The campaign is initiated either via deceptive spam emails or when a victim visits a compromised website. The victim is first served with a javascript payload which tracks them using a unique marker.
The initial payload collects fingerprinting information about the victim which is sent to the attacker’s server, the server in turn using this information to determine the next redirection. The next level of intermediate redirections use random domains before delivering a malicious payload. The final malicious payload observed is either a potentially unwanted application/program (PUA or PUP), scareware, or a notification scam.
ApateWeb has a variety of tricks up its sleeve to avoid detection and deter analysis by security researchers:
The campaign forwards traffic to the next layer only if a victim's browser retrieves an entry URL with specific parameters. Anyone browsing the website of an ApateWeb-controlled domain directly is redirected to a popular search engine or receives an empty page. This strategy helps the campaign to protect its domains from being blocked by security crawlers performing periodic scans of websites.
If a security crawler accesses the entry URL, ApateWeb tries to cloak itself by showing an error page. The campaign detects crawlers and bots by inspecting their user agent.
ApateWeb controls more than 10,000 registered domains and abuses wildcard DNS, which allows the campaign to deliver malicious content via a virtually infinite number of subdomains.
Symantec's network protection technology, Intrusion Prevention System (IPS) blocks ApateWeb redirection attempts to prevent infection/damage to the system. The attack is blocked at the initial stage, thus ensuring that no malicious payload is dropped onto the system.
Symantec protects you from this threat, identified by the following:
Network-based
Web Attack: Malicious Redirection 44
Malicious Site: Malicious Domain Request 143
Malicious Site: Malicious Domain Request 142
Malicious Site: Malicious Domain Request 164
Web Attack: Malicious JavaScript Download 55
Web-based
Observed domains/IPs are covered under security categories in all WebPulse enabled products.
Click here to learn more about managing Symantec's Endpoint Security Intrusion Prevention System (IPS).
Click here to learn more about Symantec's cloud-based Web Security Engine (WebPulse). |
Protection Highlight: Attack Group Remediation
All across the world, cyberattacks are increasing on a daily basis. Increasing in number. Increasing in sophistication. Increasing in daring. Cybersecurity vendors can't afford to take their collective eyes off the ball for a second, or the bad guys will find a way in. Unfortunately, they sometimes do get in. Which makes it all the more pressing for the security industry to get creative in coming up with ways to fight cybercriminals and protect our customer's most critical assets - their information and often their reputation.
While static detection signatures still have a part to play and probably always will, behavioral detections are without doubt the flavor of the times. Whereas a static signature is like a light switch - either on of off, i.e. the file is either good or bad - a behavioral analysis technology can determine degrees of badness. And a set of behavioral analyses combined can be even more deterministic of whether the behavior is "bad enough" to block. To add perspective, sometimes legitimate but poorly written code can behave pretty badly, but obviously shouldn't be blocked.
Symantec behavioral security technologies work to secure networks from fileless, Living Off the Land (LOTL), behavior-based attacks. These dynamic technologies monitor all relevant activity on an endpoint, learn normal application behavior patterns, and are trained to alert or quickly block deviations from the norm. Along with Behavior Analysis and System Heuristics (BASH) and Behavioral Policy Enforcement (BPE), Attack Group Remediation (AGR) is one of these technologies. AGR is a recently introduced feature that identifies all the components in a detected attack, and ensures that each process and thread that is part of the attack is removed.
Traditional malware detection blocks or remediates a piece of malware. This alone is often enough to effectively halt the attack. But not always.
Malware often utilizes complex attack chains that include multiple LOTL processes. If any of these processes is left running, the attack can restart or continue even after key components are detected and removed.
AGR applies the BASH grouping feature to place processes and files into dynamically generated groups by tracking process lineage, file lineage, and thread injection. Whenever malware is detected, its entire group gets convicted.
To convict the group, AGR determines each running process that is part of the group, and convicts the malicious processes through the use of special BPE signatures.
Every day, AGR terminates hundreds of thousands of malicious processes and breaks the attack chains of a multitude of threats, including backdoors, loaders, infostealers, ransomware, etc. Our telemetry systems have recorded well over a million blocks in the last 30 days alone.
We recently published a bulletin about Mallox ransomware, still very active targeting enterprises and organizations around the globe. There is clearly a diverse range of protections in place for this threat (blocking malware as *anything* is obviously preferable to missing it trying to get the name right) and you'll notice AGR.Terminate!g2 is one of them. To demonstrate the efficacy of this behavioral technology, our malware analysts tested some recent Mallox ransomware samples and confirmed almost all were caught by AGR (what wasn't blocked by AGR was stopped by the multitude of other detections listed). A good result, but the game doesn't stop there. The endless drive to innovate and improve must go on.
To see a list of previous bulletins which mention Attack Group Remediation, click here.
To learn how Symantec behavioral security technologies provide protection against zero-day attacks, click here.
To learn more about Symantec's Endpoint Security Service, click here. |
Protection Highlight: AveMaria RAT blocked by multiple protections
AveMaria (also known as the Warzone RAT) is a Remote Access Trojan first seen around the end of 2018 which attempts to steal information from victims, although some variants are also capable of additional functions such as remote desktop access, escalating user privileges, and activating the camera. It typically spreads via spam email "phishing" campaigns containing malicious attachments or in some cases links to malicious files hosted on legitimate cloud services and file-sharing platforms. Relative to some other RATs, AveMaria is not what you'd call highly prevalent, but it is persistent with spam runs regularly targeting a wide range of business sectors, and while it does appear to focus mainly on the EMEA region, it also includes industries in the U.S., Middle East and Asia Pacific.
Here at Security Response we get alerted to and investigate varying degrees of threat activity involving multiple threats on a daily basis, but on January 4 no less than three separate protection technologies triggered simultaneously alerting us to a significant spike in AveMaria detections. Our .NET Emulator picked it up as MSIL.Downloader!gen8, our Machine Learning technology blocked the attack as Heur.AdvML.B, and one of our Scoring Heuristic engines revved to the tune of Scr.Malcode!gdn32.
It's worth noting this particular attack employs the double extension trick, using a ".pdf.exe" executable compressed inside a .gz archive file which is attached to the email.
Symantec has 0-day protection in place, identified by the following:
File-based
MSIL.Downloader!gen8
Scr.Malcode!gdn32
Machine Learning-based
Heur.AdvML.B
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers.
To learn more about Symantec's Cloud Email Security Service, click here.
To learn more about how Symantec Endpoint Protection uses advanced machine learning, click here. |
Protection Highlight: BatCloak darkens the threat landscape
Due to their simplicity, built-in Windows functionality, concealment potential, and scripting capabilities, batch files (a script file with a .bat extension used to execute commands in serial order) have long been used by threat actors for various purposes, but primarily as loaders of additional malicious files.
Several months ago an obfuscation tool for batch files known as BatCloak came to the attention of security researchers, with multiple reports of malicious batch loaders obfuscated by it and used in campaigns around the world to delivered malicious payloads such as infostealers, remote access trojans, etc. Detections recorded by Symantec security products through September show a fairly consistent wave of attacks taking place.
Symantec recently observed an actor posing as an Indian HR company that provides staffing services, primarily to the engineering, hospitality, and retail sectors. Malicious emails were sent to companies and government agencies worldwide, disguised as staff recruitment offers. The attachment masquerades as a CV and comes in the form of a zip file with a bat file inside it (Description for your reference.zip > Description for your reference.bat).
If an unsuspecting user extracts the batch file and runs it, they will actually be executing Agent Tesla - an infamous and highly prevalent infostealer that we have posted about here many times.
Symantec protects you from this threat, identified by the following:
Behavior-based
SONAR.BatCloak!gen1
SONAR.BatCloak!gen2
File-based
Trojan.BatCloak
Trojan.BatCloak!gen1
Trojan.BatCloak!gen2
Click here to learn more about Symantec's Endpoint Security Service.
Click here to learn how Symantec behavioral security technologies provide protection against zero-day attacks. |
Protection Highlight: Beware of gift card and rewards program scams
We're right in the midst of the holiday season, and gift card and rewards program scams are running wild. This is an enduring scam that continues to be prevalently used worldwide, with increases around gift and sales seasons. These malicious gift card and rewards program scams typically fall under the umbrella of phishing attacks. They often play on our love for freebies and discounts.
Here's the lowdown: scammers send out enticing messages claiming you've won a free gift card or some exclusive reward from a well-known brand. Once someone takes the bait and clicks on the link or follows their instructions, the users might be asked to provide personal information or download a file containing malware. Some might even trick users into making a payment to claim their "free" gift. It's a classic bait-and-switch, leaving the victim with empty hands and potentially compromised sensitive info.
Symantec tracks a multitude of campaigns, and one recently caught our eye due to the number of URLs (283k) that popped up in recent weeks. Below are the numbers of URLs related to this particular campaign that we've observed.
The URLs used 294 unique, recently registered domains, which appear to be DGA-generated using names, words, and word parts. The domains used uncommon Top-Level Domains (TLDs) such as beauty, boats, bond, cfd, fun, homes, lat, monster, sbs, etc. Examples include the following:
masonblackwell[.]beauty
harmonyx[.]cfd
jayleo[.]bond
nebulashift[.]boats
eliwhitestudio[.]monster
This type of social engineering and attack usually targets consumers, but enterprise users are nonetheless not less impacted. In fact, the line between personal and professional online behavior has become increasingly blurred. Employees, while using enterprise machines, may inadvertently engage with these scams, posing a significant risk to the company's infrastructure.
Symantec has 0-day protection in place against this gift card and rewards program scams campaign, identified by the following:
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers.
Web-based
Observed domains/IPs are covered under security categories in all WebPulse enabled products. Newly observed URLs are detected and categorized using known path patterns.
Customers are advised to enable WebPulse real-time categorization, for best protection. Click here for instructions on configuring WebPulse.
Click here to learn more about WebPulse, Symantec's cloud-based Web Security engine.
Click here to learn more about Symantec's Cloud Email Security Service. |
Protection Highlight: Beyond borders – Japan faces endless localized phishing emails
On a daily basis, Japanese consumers and enterprise users are constantly bombarded by a plethora of malicious emails whose purpose is to phish for sensitive data and/or scam, leading to identity theft, financial loss, account takeover, spear phishing, and more. Over the years, the actors behind these endless waves continue to purport themselves as services that are most frequently used by Japanese internet users—often linked to financial services like credit cards.
Despite the fact that the majority of these groups and individuals are not from the local area, their malicious emails are increasingly composed in Japanese, displaying varying degrees of grammatical and spelling proficiency; the same goes for the fake websites. This increase can be attributed to the continuous improvement of translation services.
As these tools become more advanced especially with AI, cybercriminals are able to easily translate their phishing content into the target language, making it appear more authentic to potential victims. This not only helps them overcome language barriers but also allows them to tailor their messages to specific regions and demographics, making the phishing attempts more convincing. However, they still struggle with accurately translating this highly contextual language with a unique grammar structure and intricate nuances for several reasons:
Politeness and honorifics
Hiragana, katakana and kanji characters
Context and ambiguity
Cultural nuance
Syntax and word order
Lack of parallel data
Here is a snapshot of the targeted services that Symantec observed this month in the Japanese phishing threatscape.
At this time, Amazon and Aeon users are the most targeted services. However, we have also observed targeting of others, such as Apple, NTT, SMBC, AU, EPOS, ETC, Ekinet, Saison card, American Express, and Mitsukoshi Isetan, among others. Furthermore, criminals primarily abused .cn, .shop, and .cfd TLDs to create their fraudulent domains.
Examples of observed email subjects (as is - including typos):
Amazonプライム会費のお支払い方法に問題があります
えきねっとにシステムを更新する
えきねっとサービスご利用者様へ大切なお知らせ
三井住友銀行アカウントの異常通知
[AMERICAN EXPRESS] ご請求金額確定のご案内
「ETC利用照会サービス」アカウントの有効性を検証する必要がある
Apple IDの監視:非常に重要なお知らせ
三菱UFJ会社から緊急のご連絡
セゾンカード: 一瞬で終わる安全確認をお願いします!
【SMBCからの大切なお知らせ】ネットショッピングご利用時の本人認証方式変更のお願い
【SAlS0Nカード】ご利用確認のお願い
Aeon Payアカウントでの異常な取引が検出されました。
Symantec has protection in place, identified by the following:
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers.
Web-based
Observed domains/IPs are covered under security categories.
To learn more about Symantec's Cloud Email Security Service, click here.
To learn more about Symantec's cloud-based Web Security Service (WebPulse), click here. |
Protection Highlight: BlackByte Ransomware
Notorious ransomware actors have been utilizing various threat-related Tactics, Techniques, and Procedures (TTPs) for years to execute their attacks and accomplish their objectives. Among these actors, BlackByte stands out. Recent reports suggest that they can complete an attack within 5 days by leveraging an array of vulnerabilities, backdoors, and tools to infiltrate, persist, conduct reconnaissance, move laterally, exfiltrate data, and encrypt information.
ProxyShell and ProxyLogon vulnerabilities are targeted by various threats (including ransomware) as an initial means of infiltrating systems. The following chart shows recent network blocks of these attempts.
MITRE TTPs known to be used by BlackByte's actors include the following:
* Signed Binary Proxy Execution: Rundll32 [T1218.011]
* Command and Scripting Interpreter: Windows Command Shell [T1059.003]
* Boot or Logon Autostart Execution: Registry Run Keys / Startup Folder [T1547.001]
* Software Discovery: Security Software Discovery[T1518.001]
* Remote Access Software [T1219]
* Remote System Discovery [T1018]
* System Network Configuration Discovery: Internet Connection Discovery [T1016.001]
* Impair Defenses: Disable or Modify Tools [T1562.001]
* Inhibit System Recovery [T1490]
* Modify Registry [T1112]
* Domain Trust Discovery [T1482]
* System Network Configuration Discovery [T1016]
* Account Discovery [T1087]
* Permission Groups Discovery [T1069]
BlackByte ransomware initially gained infamy in 2021 when it was first observed, and over the years it has repeatedly made headlines by targeting diverse industries such as healthcare, manufacturing, and government entities. Presently, this actor remains active and has reportedly victimized multiple organizations over the past two months.
Symantec provides zero-day protection against this threat, identified by the following:
Behavior-based
ACM.Adfnd-Lnch!g1
EDR-based
Symantec EDR is capable of monitoring and flagging this threat actor's tactics, techniques and procedures.
Symantec added malware specific threat hunting queries which can be fired on the customer iCDM console. For more information on these queries, refer to this GitHub repository: https://github.com/Symantec/threathunters/tree/main/Ransomwares/BlackByte
Symantec Endpoint Detection and Response (EDR) uses machine learning and behavioral analytics to detect and expose suspicious network activities. EDR alerts about potentially harmful activity, prioritizes incidents for quick triage, and allows incident responders to navigate device activity records for purposes of forensic analysis of potential attacks.
File-based
Ransom.Blackbyte
Network-based
Audit: AnyDesk Remote Desktop Activity
Audit: ADFind Tool Activity
Attack: Microsoft Exchange Server CVE-2021-26855
Web Attack: Microsoft Exchange Server RCE CVE-2021-34473
Web Attack: Microsoft Exchange Server CVE-2021-34473
Web Attack: Microsoft Exchange Server Elevation of Privilege CVE-2021-34523
SEP's Audit Signatures are intended to raise awareness of potentially unwanted traffic in the network. By default, they do not block. Administrators reviewing the logs of IPS events in their network can note these Audit events and decide whether or not to configure the corresponding Audit Signatures to block the traffic.
Machine Learning-based
Heur.AdvML.B
Policy-based
Symantec Data Center Security (DCS) out-of-the-box default hardening policies for Windows provide 0-day protection against unknown threats, including previously unseen ransomware variants and associated behaviors.
Symantec DCS default hardening policies for Microsoft Exchange servers protect against ProxyShell vulnerabilities.
Click here to learn more about Symantec's Endpoint Security Service.
Click here to learn more about managing Symantec's Endpoint Security Intrusion Prevention System.
Click here to learn how Symantec behavioral security technologies provide protection against zero-day attacks.
Click here to learn more about how Symantec Endpoint Protection uses advanced machine learning.
Click here to learn more about Symantec's Data Center Security.
Click here to learn about Symantec Endpoint Detection and Response. |
Protection Highlight: Canadian banks and fintech SMS phishing show no signs of stopping
Symantec observes malicious SMS (also known as smishing) that lead to scams, phishing, and malware globally on a daily basis. Although these aren't new, it's increasingly happening, and many purport themselves as well-known financial institutions. One example is Canadian banks and fintechs, of which Symantec blocked multiple smishing campaigns in recent weeks.
In these campaigns, consumers are the primary targets, but enterprise users are also at risk due to the increased usage of company phones and the "Bring Your Own Device" (BYOD) trend. The impact and consequences of these smishing attacks are severe for both groups. Users who are successfully lured by these messages are redirected to phishing websites that mimic banks, with the intention of stealing sensitive information, ultimately leading to financial loss, identity theft, account takeover, emotional distress, and more.
Malicious SMS examples (as is - including typos):
RBC CANADA Security Alert : Your Client Card 4519**** has been temporarily locked due to unusual activity. To Regain Access Visit hxxps[:]//royal-banking-casecurity-verifcationonline[.]com
SCOTIABANK: As a resuIt of new onIine security measures, A temporary hold has been placed on your access card starting with 45360*. For help reactivating your card, please visit: hxxp[:]//scotian-onlines-resets[.]com
Info RBC: Vous avez recu (2) message important. Voir hxxps[:]//royalbank-messages[.]com
RBC: A temporary hold has been placed on your access card starting with 4519**. To reactivate your online access, please visit: hxxps[:]//rbc-profile[.]com
CIBC SECURITY ALERT As a resuIt of new onIine security measures, A temporary hold has been placed on your access card starting with 45064*. For help reactivating your card, please visit: hxxps[:]//cibcreset[.]info
INTERAC e-Transfer: Canada Revenue Agency has issued you a refund on 20/03/2023 after making calculation errors on your last T4 return. Remember to deposit your interac E-transfer of 583.18$ CAD at: hxxps[:]//etransfer-interacdeposit[.]com/
INTERAC e-Transfer: Canada Revenue Agency has sent you money. See hxxps[:]//secure-e-transfer-acceptonline-cra[.]com
Symantec has protection in place, identified by the following:
Mobile-based
Symantec Endpoint Protection Mobile is able to analyze links contained within SMS messages. It shields users from SMS phishing attempts by checking URLs found in text messages against the threat intelligence in Symantec WebPulse, part of the Symantec Global Intelligence Network (GIN), and alerting users when the links are suspect. The fake domains used in this campaign are known to WebPulse.
Web-based
Observed domains/IPs are covered under security categories in all WebPulse enabled products.
Additionally, to mitigate the effectiveness of smishing attacks, it's crucial for individuals to stay vigilant, be cautious of unsolicited messages, avoid clicking on suspicious links, and verify any communication directly with their bank using official contact details.
To learn more about Symantec Mobile Endpoint Protection, click here.
To learn more about Symantec's cloud-based Web Security Service (WebPulse), click here.
To learn more about Symantec's Global Intelligence Network (GIN), click here. |
Protection Highlight: Chaos ransomware still causing chaos
We've posted over a dozen bulletins directly or indirectly relating to Chaos ransomware during the past year or so, an indication of it's prevalence in the threat landscape. Chaos ransomware first appeared in 2019 and has been responsible for numerous attacks on businesses and individuals worldwide. With so many variants in play now the malware is delivered in a wide variety of different ways but still typically via phishing emails or malicious websites, aiming to exploit vulnerabilities in outdated software in order to gain access to the victim's system. Once installed, Chaos encrypts the victim's files and displays a ransom note demanding payment in exchange for the decryption key.
One reason why there are so many variants of Chaos ransomware in the threat landscape is that the source code of the malware was leaked online in 2020. This made it easier for cybercriminals to create new strains of the ransomware, customize them for their own agenda and use them to carry out attacks. Multiple groups and individuals worldwide currently employ Chaos ransomware, targeting both consumers and organizations of various sizes and sectors, with single or multiple machine infections.
Another reason for the proliferation of Chaos variants is that the use of ransomware in general is a lucrative business. Ransomware attacks can generate substantial profits, especially if the victim is a sizeable organization willing to pay a large ransom to recover their data. The success of previous Chaos attacks has likely encouraged other threat actors to create their own versions of the ransomware. All the more reason to protect your network large or small with a reputable security solution which has multiple protective layers, optimally configured to ensure your security posture is rock solid.
While the attacks are fairly constant over time, February 12~14 saw a spike due to yet another new variant.
Symantec has 0-day protection in place, identified by the following:
Behavior-based
SONAR.*
Machine Learning-based
Heur.AdvML.*
* This indicates multiple similarly named detections exist such as SONAR.Heur.Dropper, SONAR.Psdownloader!g1; Heur.AdvML.B, Heur.AdvML.C, etc.
To learn more about how Symantec Endpoint Protection uses advanced machine learning, click here.
To learn how Symantec behavioral security technologies provide protection against zero-day attacks, click here. |
Protection Highlight: Cloudflare R2 Phishing Abuse
Phishing actors, whether individuals or groups, have long used various methods to deliver their phishing campaigns. Content Storage Networks (CDNs) such as IPFS and CloudFlare R2 are among the most abused to host phishing pages. We recently published a protection bulletin about a single campaign, but over the past 30 days we have observed many more instances worldwide, primarily attempting to steal email credentials from enterprise users.
One can assume the main reasons phishing actors increasingly exploit CloudFlare R2 are due to its free tier, ease of use, strong reputation, and global reach. All these factors combined provide them with a robust platform to conduct their social engineering tactics, aiming to achieve a certain level of evasion, resilience, and anonymity.
As mentioned, most observed campaigns target email credentials by enticing users with emails related to account issues (password problems, termination, unread emails, etc.), bills, police warnings, and other social engineering tactics. The emails contain a malicious URL that directs to a phishing page, with the user's email address at the end of the URL. Consequently, if users are successfully tricked into clicking the URL, the fake login page will already display their email address in the login field making the process seem that much more credible.
Below are some recent examples of malicious URLs that lead to phishing pages hosted on CloudFlare R2:
hxxp[:]//pub-733372c603ef451496fbd54cfcb41576[.]r2[.]dev/93306DHI[.]html#[victim's email address]
hxxp[:]//pub-be898b69352444c28d68f43e8725f2d1[.]r2[.]dev/godisalive[.]html#[victim's email address]
hxxp[:]//pub-f4d1302dafbf4beeaf3e5e773e67edc4[.]r2[.]dev/allupdate[.]html#[victim's email address]
hxxp[:]//pub-ad5b0662c2a54e5884a831384bd99913[.]r2[.]dev/pagefem345[.]html#[victim's email address]
hxxp[:]//pub-ad60cadbed8e448499578f472c0a3183[.]r2[.]dev/af[.]html#[victim's email address]
hxxp[:]//pub-a8906372f15e4c3c9eeedea91a48a923[.]r2[.]dev/index[.]html#[victim's email address]
Symantec protects you from this threat, identified by the following:
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers.
Web-based
Observed domains/IPs are covered under security categories in all WebPulse enabled products. |
Protection Highlight: Cryptocurrency mining attacks continue
Crypto miners have been a prevalent presence in the digital landscape ever since Bitcoin was introduced, the pioneering cryptocurrency. Over the years, as the popularity of digital currencies soared and the market expanded, crypto mining activities also experienced steady growth. Many individuals and organizations see crypto mining as a promising investment opportunity, utilizing their computing power to mine cryptocurrencies and potentially reap substantial profits.
However, with the increasing mainstream adoption of cryptocurrencies, the field attracted the attention of cybercriminals as well. Unbeknownst to many consumers and enterprises, their computers and infrastructure became prime targets for crypto mining attacks.
Cybercriminals behind these malicious crypto currency miners have devised various sophisticated techniques to infect systems and compromise computational resources secretly. They utilize techniques like malicious emails, compromised websites, drive-by-download and vulnerabilities to gain unauthorized access to machines (as well as persists and move), subsequently turning them into unsuspecting miners.
These actors have been using both custom miner and legitimate ones to fuel their campaigns. Despite this, potential misuse of the legitimate mining tools, Symantec has had long standing detection of it. Symantec observes tens of thousands cryptocurrency miners monthly.
The impact of these illicit mining operations are significant. While consumers experience sluggish computers, encounter unexpected spikes in electricity bills, and notice reduced battery life on their devices. For enterprises, the consequences are more severe, involving disruptions to critical operations and potential data breaches.
Symantec has had long-standing detection of crypto miners and strong 0-day protections are in place:
Behavior-based
SONAR.Bluwimps*
SONAR.CoinMiner*
SONAR.Coinbitminer!g1
SONAR.GhostMiner!gen1
SONAR.Gosopad!gen5
SONAR.Miner*
SONAR.Suspdrop!g61
File-based
CL.XMRig!gen1
Linux.Coinminer
MSH.Gosopad
Miner.Bitcoinminer
Miner.Burst
Miner.Cpuminer
Miner.Jswebcoin*
Miner.Neoscrypt
Miner.Wasmwebcoin
Miner.XMRig*
Miner.Zcash!gen1
Miner.Zcashminer
OSX.Coinminer
OSX.Miner.Coinminer
PUA.Bitcoinminer
Trojan.Adylkuzz
Trojan.Coinbitminer
Trojan.Coinminer*
Trojan.Madominer
Trojan.Minjen*
Trojan.Shminer
W32.Coinbitminer
W32.Mysracoin
W32.Rarogminer
W32.Rarogminer!G1
W32.XiaobaMiner
Machine learning-based
Heur.AdvML.*
To learn how Symantec behavioral security technologies provide protection against zero-day attacks, click here.
To learn more about Symantec's Endpoint Security Service, click here.
To learn more about how Symantec Endpoint Protection uses advanced machine learning, click here.
To learn more about Symantec's Data Center Security, click here. |
Protection Highlight: Cryptojacking
Browser-based mining is a method of cryptocurrency mining that happens inside a browser using scripting language. This method differs from the more commonly known file-based cryptocurrency mining approach, which requires downloading and running a dedicated executable file. If a web page has a coin-mining script injected into it, visitors to the web page will have their computing power used to mine for cryptocurrency for as long as they keep the web page open. Some websites may use coin mining as an alternative to advertising to generate revenue, which is fine provided customers are told that their CPU power will be used to mine cryptocurrency while they are visiting that website. However, some websites mine cryptocurrency using the website visitor’s computer or mobile resources without their knowledge or consent. This is known as cryptojacking via browser-based coin mining. Symantec blocks thousands of these attacks every day.
In a typical browser-based cryptojacking scenario, an attacker compromises a website and injects a few lines of JavaScript code into the website code. This injected code is responsible for mining cryptocurrency for the attacker at the behest of website visitors. Monero is an example of a cryptocurrency that enables browser-based mining. The cryptocurrency uses the RandomX hash function, a hashing algorithm used for certain PoW blockchains, such as Monero. Historically, cryptojackers either leveraged JavaScript-based cryptojacking services, such as Coinhive, which gained attention in 2017, or relied on compromised plugins or malicious browser extensions to deliver malicious JavaScript.
Unlike threats such as ransomware which immediately disrupt victims’ access to their devices, cryptojacking could be quietly carried out on a victim’s device for a long time before they realize what's happening. Even fully-patched devices can be targeted via browser-based coin miners. The primary impact of such cryptojacking is performance-related. Potential impacts include slowdown in device performance, overheating batteries, devices becoming unusable, and higher costs due to increased electricity usage for businesses operating in the cloud that are billed based on CPU usage. Additionally, loss of customers and goodwill are also potential impacts for website owners and businesses.
Some cryptojacking examples include:
* Cryptojacking Operation Targeting Kubernetes Clusters for Dero Mining in Feb 2023.
* RapperBot DDoS malware adds cryptojacking as new revenue stream.
* Internet-exposed Linux and Internet of Things (IoT) devices are being hijacked in brute-force attacks as part of a recently observed cryptojacking campaign.
Symantec's Intrusion Prevention System (IPS) technology protects customers against browser-based cryptojacking by blocking associated malicious network activity using a variety of detections, identified by the following:
Network-based
Web Attack: JSCoinminer Download*
Web Attack: JSCoinminer Website*
* This indicates multiple similarly named detections exist such as Web Attack: JSCoinminer Download 1, Web Attack: JSCoinminer Download 2, etc.
To learn more about managing Symantec's Endpoint Security Intrusion Prevention System (IPS), click here. |
Protection Highlight: DarkGate goes PDF
We posted about DarkGate - a Windows-based malware with a wide-range of capabilities including information stealing and remote access - in August, spreading via malicious ads and SEO poisoning, and mentioned it again just last month in relation to a former affiliate of Qakbot known as TA577 who was observed delivering DarkGate via its malicious email campaigns in which it notably employed an extensive array of attack chain combinations.
Even more recently, DarkGate has turned to leveraging PDF attachments. In some cases, the filename consists of only one or two alphabet characters, along with a malicious link that gets scanned within the file's contents. In these cases, the link downloads a ZIP file which contains either a malicious VBS or MSI file, with a second-stage payload being an AutoIT scripting that holds the actual DarkGate malware code.
A third PDF-related infection was also observed. With these, the link inside the attached PDF file (most having much longer file names) downloads a URL shortcut which in turn connects to a WebDav server controlled by the attacker in order to execute either a malicious VBS, JS or MSI file.
PDF > URL > .URL shortcut > MSI/JS/VBS > AutoIT > DarkGate
Our telemetry systems reported this latest campaign starting in earnest on October 13, with a sizeable spike on October 25 followed by a smaller one on October 27.
Symantec protects you from this threat, identified by the following:
File-based
Scr.DLHeur!gen1
Scr.DLHeur!gen6
Scr.DLHeur!gen7
Trojan.DLHeur!gen5
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers.
Web-based
Observed domains/IPs are covered under security categories in all WebPulse enabled products.
Click here to learn more about Symantec's Endpoint Security Service.
Click here to learn more about Symantec's Cloud Email Security Service.
Click here to learn more about Symantec's cloud-based Web Security Service (WebPulse). |
Protection Highlight: Dark Power Ransomware
Symantec constantly monitors ransomware actors, and one such actor which goes by the name of Dark Power has recently escalated its activities, targeting organizations and enterprises in several countries across the globe. This relatively new group employs double-extortion tactics by posting a list of victims on an onion website (these are "dark web" sites that use special software to encrypt their connections and enable anonymous communication in order to hide their location and various other identifiers) and threatening to sell stolen data if their ransom demands are not met. If encryption is successful on compromised machines files are appended with a .dark_power extension. If this criminal group does manage to achieve a degree of success, there is a high likelihood that their activities will continue and possibly intensify.
Dark Power is currently claiming victims in the following countries:
* Algeria
* Czech Republic
* Egypt
* France
* Israel
* Peru
* Turkey
* USA
Symantec has 0-day protection in place, identified by the following:
Behavior-based
SONAR.Ransomware!g1
Machine-Learning-based
Heur.AdvML.B
To learn more about Symantec's Endpoint Security Service, click here.
To learn more about how Symantec Endpoint Protection uses advanced machine learning, click here. |
Protection Highlight: Data Center Security - Ransomware protection for Microsoft SQL Server
Data Center ransomware attacks on the rise
Ransomware threats originate in many different ways and utilize various points of entry to target enterprises, but one method increasingly used by threat actors goes directly after data center servers and workloads as the initial step in the attack chain. These systems may not keep up with the recommended patches, often run legacy applications where vendor security patches may not be available, or the workloads may not be scheduled for patch updates to maintain business continuity from any patch downtime. As a result the risk exposure of a data center environment to cyber attacks and ransomware activities is very high. Breaching a single data center can provide access to multiple interconnected systems and applications, maximizing the potential impact of an attack.
Database workloads host sensitive data for an organization, and power mission critical business services and thus are valuable targets for ransomware actors to steal data and extort a ransom by encrypting critical data files. While there are many different database applications, Microsoft SQL Server is one of the most popular databases deployed globally and an irresistable target for ransomware, primarily because it is deployed on Windows where attackers have abundant malware tools at their disposal to be used as payloads, as well some that can be leveraged by living off the land.
Microsoft SQL Server - a prime target
Recent reports highlight a pattern where ransomware actors target Microsoft SQL Server as their initial entry point into data centers. Poorly configured SQL servers and weak admin passwords provide avenues for brute force attacks or SQL injection enabling unauthorized access and data exfiltration. Compromised systems may then be used as access points to be sold to other parties or used for installing additional malicious payloads ultimately for data exfiltration or financial extortion. The recent CISA directive for software publishers to take preventative measures in order to eliminate SQL injection vulnerabilities underscores the need for enterprises to pay particular attention to the security of Microsoft SQL servers. The CISA directive is in response to the SQL injection vulnerability in the file transfer application MOVEit, which has been exploited by CL0P ransomware for remote code execution.
Some of the notable cyber threat activity seen against Microsoft SQL Server includes:
Mimic ransomware where the initial access was obtained by brute force on exposed Microsoft SQL servers
Mallox ransomware where the initial access attempts were done using dictionary brute force attack, followed by cmd shell execution for further activities
CLR SQLShell, similar to the xp_cmdshell stored procedure used to execute shell commands on Microsoft SQL servers
CL0P ransomware exploited a SQL injection zero-day vulnerability CVE-2023-34362 in MOVEit file transfer application to install a web shell named LEMURLOOT
Freeworld ransomware, new variant of Mimic, also accessed by brute force on unsecure Microsoft SQL servers
Bluesky ransomware also gained initial access from bruteforce login to the sa account and then enabled the xp_cmdshell stored procedure to execute shell commands
Data Center Security - an effective solution
Symantec Data Center Security (DCS) offers a comprehensive defense-in-depth approach in securing and safeguarding Microsoft SQL servers and the underlying operating systems. Our solution has been effective in providing zero-day protection against the increasing prevalence of ransomware attacks and other cyber threats targeting data center environments.
DCS Microsoft SQL Server Sandbox
Symantec sym_win_hardened_sbp policy for Microsoft SQL Server has a built-in DCS sandbox (mssqlsrv_ps) built upon the least privilege principle to lockdown Microsoft SQL workload:
Network controls define the perimeter for client applications to trusted networks and devices thereby restricting initial access. Additionally you can allow network connections only on acceptable ports
Software Execution control when enabled prevents unauthorized and arbitrary execution of command interpreters such as cmd.exe, cscript.exe. This is effective to prevent living off the land techniques to run system commands for malicious activities and reaching out to C2 servers for downloading payloads and further commands
Software Install Restrictions and Operating System Restrictions prevent any attempts to maintain persistence for subsequent access and tampering of critical Windows operating system resources
Process Access Control prevents dumping of LSASS using procdump or Mimikatz tool
Protected App Control ensures that while MS SQL processes are trusted to execute, they cannot be exploited to install malware payloads in the form of batch files, powershell scripts, tools such as Cobaltstrike, Mimikatz and ransomware utilities
Symantec Windows Baseline Detection Policy
Symantec Windows Baseline Detection policy has a Microsoft SQL Server Monitor ruleset to provide visibility into SQL server events and alert on possible suspicious activities:
Microsoft SQL Server Login Activity Monitor audits successful and failed logins to the Microsoft SQL server sa account, useful to alert on brute force attacks
Microsoft SQL Server Service Activity Monitor reports start and stop of Microsoft SQL services
Microsoft SQL Server File and Registry Monitor provides real time monitoring of any tampering with the SQL server files and registry resources
Safeguarding data center environments, particularly database workloads, is paramount. Enterprises must prioritize security measures, including timely patching, robust access controls, and ongoing monitoring in order to mitigate the risk of ransomware infiltration and protect sensitive data from exploitation. These Symantec Data Center Security default policies are enabled out-of-the-box to ensure robust protection against the rising trend of entry point attempts against data center servers and workloads.
Click here to learn more about Symantec's Data Center Security. |
Protection Highlight: DCS 0-day protection for Windows Servers against Royal ransomware
We have posted several bulletins recently about a relatively new ransomware family called "Royal". Royal spreads through various means such as callback phishing, malware loaders such as Batloader and Qbot, application vulnerabilities (CVE-2022-27510 is a recent one) as well as a variety of widely available open-source tools. It has also been reported to target virtual machines. Malware loaders are programs that establish communication between the attacker and the target system, typically representing the initial stage of an attack. These early stage loaders spread using common methods such as malicious ads, email spam containing malicious links or embedded files, fake websites, forums, etc.
Royal encrypts files using a .royal file extension, drops a readme.txt file directing victims to a Tor-based payment site, and deletes backups and volume shadow copies in order to add leverage to it's ransom demands. It can also encrypt network shares, employing multi-threaded execution in order to accelerate the encryption process. The group behind Royal targets multiple business sectors including healthcare, insurance, industrial companies and even attacked a popular UK racing circuit.
As our previous posts on Royal demonstrate, Symantec has detections for this ransomware in several of our protection technologies, including static file detections, behavioral detections, machine learning heuristics and network-based signatures, however, with our observations on the Royal ransomware targeting Windows servers on the network, we also want to highlight Symantec Data Center Security (DCS), our server security solution.
Symantec Data Center Security out-of-the-box default hardening policies for Windows provide 0-day protection against unknown threats, including previously unseen Royal ransomware variants and behaviors as follows:
Policy-based
DCS Software Install restrictions prevent ingress transfer of malware tools as well as remote installation and execution of Royal ransomware using psexec.
Process launch restrictions prevent arbitrary system command execution and misuse of dual use tools for malicious activities.
DCS Sandboxing prevents tampering of critical system files and registry resources.
For additional protection, customers can configure DCS network rules to set up a network perimeter for privileged services and applications.
To learn more about Symantec's Data Center Security, click here. |
Protection Highlight: Download Insight - Zero-day protection through File Reputation
Untrusted downloads open the door to a myriad of online threats. Malware, ransomware, and even sophisticated cyberattacks can infiltrate your network through a single infected file. Once inside, these threats can wreak havoc, stealing sensitive data, disrupting operations, and causing financial loss. One such entry vector lies in "portals" – applications like browsers, downloaders, chat clients and archivers that act as gateways for files to enter your system. Like a security guard stationed outside your front door, Symantec Download Insight keeps a watchful eye on multiple portals where files get downloaded in real-time.
What sets Download Insight apart is its ability to protect against even the most novel threats using Symantec File Reputation. Symantec File Reputation goes beyond simply identifying threats, it understands the context of each file within the global threat landscape. Files are rated as suspicious when they are correlated with traits such as bad user-machine behavior, recent age, low prevalence, untrusted signers, untrusted sources and hundreds of other indicators of poor reputation. Combining these tell-tale markers of poor reputation with the fact that the file was introduced through a portal results in a block.
Symantec File Reputation includes reputation for over 9 billion files, continuously updated based on live telemetry from millions of endpoints, multiple threat feeds, and expert analysis from a dedicated team of Symantec Security Response analysts.
Behavioral Blocks - Zero-day protection with Behavioral Engine and File Reputation
Symantec File Reputation is also used in combination with the Symantec Behavioral Protection Engine, known as SONAR. SONAR tracks suspicious process behaviors such as droppers, tampering, ransomware behaviors, etc. and uses Symantec File Reputation to corroborate these suspicious behaviors with poor reputation.
Some high-profile examples of new threats proactively blocked with SONAR and File Reputation include:
Ransomware
Conti
Quasar
Sality
Emotet
Qakbot
Trickbot
Mimikatz
Wannacry
Click here to learn more about Symantec Download Insight.
Click here to learn more about how Symantec Endpoint Protection uses Symantec Download Insight to make decisions about files.
Click here to learn how Symantec behavioral security technologies provide protection against zero-day attacks. |
Protection Highlight: EDR vs Defense Evasion
Defense Evasion
Cybersecurity has been likened to an endless game of whack-a-mole. For every instance of malware that gets smacked down, another two seem to pop up. Threat actors are evolving, adapting, and developing evermore sophisticated attack strategies. Defense Evasion is one of the crucial elements in any attack chain. In order to avoid detection and maintain persistence on penetrated systems, attackers (often referred to as "adversaries") use various techniques such as disabling security products, abusing and leveraging trusted processes, masquerading, and many other tactics during an attack.
As per MITRE ATT&CK guidelines, Defense Evasion is "The adversary is trying to avoid being detected". Nice simple definition. However Defense Evasion is anything but simple, consisting of dozens of techniques, each describing a particular method or approach an attacker may use to avoid detection while attempting to compromise a target system or network (for reference, MITRE ATT&CK techniques are part of the MITRE ATT&CK framework that categorizes the methods and tactics employed by threat actors during various stages of a cyber attack. There are hundreds of different attacks, grouped under a handful of categories called "tactics" - Defense Evasion is one of these categories).
Endpoint Detection & Response (EDR)
In order to combat these relentless attacks, Symantec is continually adding new capabilities to its EDR portfolio, investing in advanced levels of protection. These new tools are available as features in our flagship Symantec Endpoint Security Complete (SESC) product, designed to address security issues and the MITRE attack chain early in the cycle to quickly detect breaches and disable attacks in progress. EDR is then able to show attack lineage through the EDR Incidents generated in the process of responding to the attacks.
Some of the most-used MITRE Defense Evasion techniques include the following:
Impair Defenses [MITRE : T1562]
This is one of the top techniques used by adversaries to disable security products, either by killing processes or uninstalling/disabling them by modifying registry entries.
Top methods
Kill process using taskkill
Disable firewall using netsh
netsh firewall set opmode disable
netsh advfirewall set currentprofile state off
Disabling security software by manipulating relevant registry entries
Symantec EDR visibility against Impair Defense attempts made by adversaries
Indicator Removal [MITRE : T1070]
Adversaries use this technique to hinder defenses by deleting or modifying artifacts generated within systems such as deleting arbitrary registry keys/values, clearing browser history, log modifications, etc.
Top methods
Clear windows events logs using wevtutil.exe
wevtutil cl security
wevtutil cl system
wevtutil cl application
Delete shadow copies using vssadmin.exe
vssadmin delete shadows /All /Quiet
Abuse Elevation Control Mechanism [MITRE : T1548]
Adversaries use this technique to circumvent system mechanisms designed to control privilege escalation by abusing permission configurations, bypass UAC, sudo caching, etc. to gain higher-level permissions.
Top methods
Bypass UAC using mmc.exe, fodhelper.exe
Symantec EDR visibility against Indicator Removal and Abuse Elevation attempts made by adversaries
To add real-world perspective, ransomware, arguably one of the most disruptive and damaging cyberthreats ever created, has multiple variants that use similar Defense Evasion techniques to those we discussed above including:
Lockbit
Enmity
Snatch
Noberus
Blackbyte
Hive
Symantec Endpoint Security Complete addresses these MITRE ATT&CK techniques and many more, all built into a single agent and a single platform. From prevention to detection to augmenting your own security resources in this non-stop game of whack-a-mole, Symantec never stops innovating to help defend your network.
Click here to learn about Symantec Endpoint Detection & Response.
Click here to learn about Symantec Endpoint Security Complete. |
Protection Highlight: Emotet 0-day protection - blocking the new with the old
The Emotet banking trojan, first discovered by security researchers way back in 2014, has turned out to be one of the most prolific and persistent trojans in the threat landscape. Originally a simple banking malware that attempted to steal sensitive banking-related information from victims, later versions saw the gradual introduction of new modular functionality and infection vectors including email spamming, malware delivery services (including other banking trojans and ransomware) and stealth modifications. The Department of Homeland Security went as far as to state that Emotet is one of the most costly and destructive malware they have ever seen.
Emotet has seen several quiet periods where activity dropped off significantly, presumably when the authorities were hot on the author's heels or perhaps due to infighting or structural reorganization. In early November however, the actors behind Emotet broke their latest hiatus by launching a new malicious email campaign complete with a new URL format and personalized landing pages. In this particular instance, the emails contain an Excel attachment which is responsible for downloading Emotet, the Excel file making use of social engineering which requires user input to successfully execute the malicious content, effectively having the unsuspecting user infect themselves.
While there are many different means of defending against malware like Emotet (as called out in our Nov 2 Emotet bulletin) the best way of course is to block it proactively, without the need to create new signatures, definitions, rules, etc., which then require their corresponding security products to be updated in order to be effective. And while there surely isn't a cybersecurity vendor in existence who would be so bold as to claim a proactive defense every time a threat comes knocking, in this case heuristic protections created months ago using multiple traits of past variants, advanced clustering technology, and a whole lot of hard work on the part of our malware analysis teams, Symantec products blocked these new, previously unseen Emotet variants without the need for any updates.
Symantec has 0-day protection in place, identified by the following:
File-based
XLM.Downloader!gen1
XLM.Downloader!gen2
XLM.Downloader!gen4
CL.Suspexec!gen128 |
Protection Highlight: FakeCop Mobile Infostealer Visits Japan
FakeCop is an infostealer which targets mobile devices, collecting various types of data including device information, contact lists and SMS messages. Once collected, the data is forwarded to the attacker's C&C servers. The authors employ XOR encryption to try to evade static detection methods but it's no match for SEP mobile's sophisticated conviction technology.
Over the past few years, FakeCop has persistently plagued mobile users in Japan with recurrent waves. The actors behind this threat haven't changed their modus operandi, continuing to leverage malicious SMS to spread their app using social engineering lures and abusing the brands of well-known Japanese telecom companies in the process.
Symantec recently observed another wave of FakeCop, again masquerading as a Japanese telecom company mobile security app ([company name]2024.apk). If users are successfully tricked into deploying the malicious app on their Android device, it will ask the user to grant several permissions needed to perform its tasks.
It will then attempt to change the default SMS app and prompt users to uninstall specific antivirus apps, the names of which are hardcoded into its program code.
Successful compromise will likely see the malicious app being spread to the victim's contacts, and may also result in the stolen data being sold on the black market potentially leading to financial loss and possibly identity theft. It could also be leveraged to conduct targeted attacks.
Symantec protects you from this threat, identified by the following:
Mobile-based
Android.Reputation.1
Click here to learn more about Symantec Mobile Endpoint Protection. |
Protection Highlight: Formbook fail, EMAIL protection win
It's hard to believe decades after its first regular use, electronic mail - or "email" as it's more commonly referred to - is still so widely used as a means of communicating, particularly for businesses. We can see the popularity of email with hundreds of billions of emails sent around the world on a daily basis and as a result it has also become a popular target for spammers. Depending on your source, it is estimated that anywhere from roughly half to well over three-quarters of all email sent today is known to be spam. A vast majority of these spam messages are “malspam”, which are spam email messages that contain or deliver malware.
Enter Formbook, one of the most infamous infostealing malware to darken our ether skies. A quick summary of Formbook's uninspiring resume:
* Formbook uses email as it's main infection vector, but may also use drive-by-download, exploit kits and software vulnerabilities
* Employs a dizzying array of email subjects but seems to prefer fairly run-of-the-mill "payment" type social engineering tactics including fake order emails, shipping and SWIFT quotes
* Around since at least 2016, consistently one of the most-seen information stealers and form-grabbers
* Primary purpose of harvesting credentials from web browsers and collecting screenshots and log keystrokes
* Has the ability to download and execute additional malicious files
* Responsible for a huge volume of credential dumps on the darknet
* Sold as Malware as a Service (MaaS) and used by multiple groups and individuals the world over
* Deploys both targeted and non-targeted campaigns against many different industries and business sectors, as well as multiple regions and countries across the globe
A real charmer. So what's the good news?
An analysis of our in-field telemetry data for the past few weeks reveals:
Over ten thousand malicious emails related to Formbook
Multiple actors and campaigns targeting multiple regions and industries
Dozens of campaigns using hundreds of email subjects
All proactively blocked by the Malware Scanning components of the ESS service.
Symantec email customers rest assured, we're in this fight to win, on your behalf.
To learn more about Symantec's Cloud Email Security Service, click here. |
Protection Highlight: Formbook gets machined
Or to be more precise, machine learned. We discussed Formbook at length in our December 8, 2022 bulletin and decided now is a good time to revisit and provide an update. To briefly recap, Formbook has been stealing information from compromised computers since around 2016, using email as it's main vector of infection along with a wide variety of subjects which include the usual fake orders, shipping invoices and SWIFT quotes. It's primary purpose is to harvest credentials from web browsers, collect screenshots and log keystrokes. Sold as Malware as a Service (MaaS), it employs both targeted and non-targeted campaigns against a range of different industries and business sectors covering multiple countries and regions around the world. Busy, to say the least. This busy in fact.
Here's an example of a typical Formbook email, this one in Bulgarian containing an attachment with a filename that ends in "pdf.zip". The fact that the email address listed in the From field differs entirely from the company name and website in the sender's signature should be a red flag in and of itself.
In this particular example, the zip archive contains an executable file which has been heavily obfuscated using a .NET packer. Once executed it copies itself to "%AppData%\Roaming\fLgUSgFvp.exe", adds an exclusion for itself to Windows Defender, creates a scheduled task for itself, and finally decrypts the actual payload injecting Formbook version 4.1 into the following Windows application: "C:\Windows\Microsoft.NET\Framework\v4.0.30319\RegSvcs.exe".
In the December bulletin we reported that the Formbook campaigns at the time had been proactively blocked by the Malware Scanning components of the ESS service, and so are the new campaigns. Several different protection technologies block Formbook, but in a testament to the effectiveness of machine learning we're happy to report in this bulletin that the logic which detects and blocks Formbook has not needed to be updated since the previous campaigns in December.
Symantec has multiple 0-day protections in place against this threat:
File-based
MSIL.Packed31
Scr.Malcode!gdn32
Scr.Malcode!gdn34
Machine Learning-based
Heur.AdvML.A!300
Heur.AdvML.A!400
Heur.AdvML.B!100
Heur.AdvML.B!200
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers.
To learn more about how Symantec Endpoint Protection uses advanced machine learning, click here.
To learn more about Symantec's Cloud Email Security Service, click here. |
Protection Highlight: Formbook sighted in the Balkans
Formbook malware is prevalent worldwide, with campaigns being observed daily and showing no signs of slowing in it's use by multiple groups and individuals. More recently, Symantec has detected this aging yet ever-present threat making the rounds in the Balkans. The actor behind this Balkan campaign is attempting to lure companies with malicious emails (subject: нов налог за купување), purporting to be from a Macedonian manufacturer of household electrical products. Attached to these emails is a malicious ZIP file containing Formbook binaries disguised as a purchase order PDF.
Active since at least 2016 and well-known in the cybersecurity world, Formbook is an infostealer that is actively sold on underground forums. It is capable of grabbing a wide range of data from infected systems including Web browser data (stored logins, cookies, form data, etc.), keystrokes, screenshots, and files stored on the computer. Below is a snapshot of recent Formbook detections.
Symantec has multiple 0-day protections in place against this threat:
File-based
Scr.Malcode!gdn34
Behavior-based
SONAR.SuspLaunch!g310
SONAR.SuspBeh!gen752
Machine Learning-based
Heur.AdvML.B
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers.
Click here to learn more about Symantec's Endpoint Security Service.
Click here to learn how Symantec behavioral security technologies provide protection against zero-day attacks.
Click here to learn more about how Symantec Endpoint Protection uses advanced machine learning.
Click here to learn more about Symantec's Cloud Email Security Service. |
Protection Highlight: GuLoader hides in plain sight
GuLoader is an advanced shellcode-based downloader which uses a variety of anti-analysis tricks to evade detection and make reverse-engineering more difficult. It has the end-goal of delivering a range of malware which may include ransomware, infostealers, banking Trojans, RATs, Proxies, etc. The Symantec Threat Hunter Team recently blogged about the use of GuLoader in the initial stages of the Bluebottle campaign targeting banks in French-speaking African countries.
We regularly observe GuLoader spreading via email spam campaigns, including versions with attached ISO files containing an embedded VBE script where the VBE script attempts to construct a PowerShell script and execute it.
Multiple GuLoader detections exist but only Scr.Malcode!gen36 are shown here
While none of this is new for GuLoader, what is interesting is the extent the authors go to in order to stay under the radar. The ISO/VBE attack mentioned above starts something like this... once the (initially obfuscated) PowerShell script is deobfuscated and the code extracted it downloads the GuLoader shellcode (in base64 format) from a directory hosted on Google drive. The GuLoader shellcode (base64 decoded) is then invoked via a Windows API (CallWindowProcA) and here the efforts at preventing or at least slowing down the analysis really kick in. These bulletins are not intended to be overly detailed or technical so we'll try to keep it simple.
GuLoader anti-analysis tricks
Anti-AV #1 : First, an attempt to prevent the operating system generating an executable from GuLoader-related shellcode blobs, a very old trick indeed.
Anti-Debugging #1 : GuLoader sets up a vectored exception handler (VEH) via a Windows API to redirect the flow of code execution to a managed exception handler which tries to catch exceptions caused by EXCEPTION_SINGLE_STEP (TrapFlag). A trap flag permits operation of a processor in single-step mode and can be manipulated in order to thwart tracers.
Anti-Debugging #2 : Anti Hardware Breakpoint and Software Breakpoint. Hardware and Software breakpoints are events whose execution will cause a halt during debugging. GuLoader checks for these and attempts to prevent them.
Anti-VM #1 : Memory Page Scanning. GuLoader uses the NtQueryVirtualMemory API to scan the entire memory of the process to check if there are any Virtual Machine (VM) or Debugging tool-related strings present.
Defense Evasion #1 : Heaven's Gate is a method introduced in the mid-2000's for compatibility purposes to allow the execution of 64-bit code inside a 32-bit process. GuLoader shellcode uses Heaven's Gate to execute a 64-bit code stub without being noticed.
Anti-VM #2 : GuLoader checks for the existence of C:\Program Files\Qemu-ga\qemu-ga.exe and C:\Program Files\qga\qga.exe, files related to the QEMU (VM) emulator.
Anti-Debugging #3 : GuLoader patches DbgBreakPoint and DbgUiRemoteBreakin to prevent debuggers from attaching to its host process.
Defense Evasion #2 : Remove Hooks in NTDLL32. GuLoader shellcode scans for SYSCALL patterns inside NTDLL, extracts the SYSCALL number and restores the function code to its original state.
Anti-Sandbox #1 : Enumerate Windows. GuLoader calls the EnumWindows API to count the top-level windows (both visible and invisible) running on the victim machine. If the number is lower than 12 the shellcode terminates.
Anti-Debugging #4 : ThreadHideFromDebugger. A common anti-debugging technique is to make use of the NtSetInformationThread API to effectively mark a thread as invisible to the debugger.
Anti-Sandbox #2 : GuLoader uses various APIs to enumerate Windows drivers, installed software and services, comparing their hashed names to pre-hashed values.
Anti-Debugging #5 : ProcessDebugPort. Guloader calls NtQueryInformationProcess to detect if a debugger is attached to its process.
Following these numerous attempts to hide, GuLoader shellcode injects itself into another process via Process Hollowing, a code injection technique in which the executable section of a legitimate process in memory is replaced (or appended to) with malicious code.
Finally, after that concerted (and obviously in our case, unsuccessful) effort at remaining undetected, GuLoader will download the final payload. In this particular case that payload is the infamous Agent Tesla.
Symantec has 0-day protection in place, identified by the following:
File-based
Scr.Malcode!gen36
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers.
To learn more about Symantec's Endpoint Security Service, click here.
To learn more about Symantec's Cloud Email Security Service, click here. |
Protection Highlight: GuLoader Revisited
We published a Protection Highlight bulletin about GuLoader back in March of this year and after observing a spike in activity last Friday, this is a short follow-up to confirm that it's still around trying to do it's dirty-work, and still successfully being proactively blocked by Symantec protection-enabled products.
To recap, GuLoader is an advanced shellcode-based downloader which uses a variety of anti-analysis tricks to try to evade detection and make reverse-engineering more difficult, with the goal of delivering a range of malware including but not limited to ransomware, infostealers, banking Trojans, RATs and Proxies. GuLoader mainly spreads via email spam campaigns with various themes, including versions with links to other malware or in some cases directly attached files containing an embedded script where the script attempts to construct a PowerShell script and execute it in order to run the final payload on the target system. The final payload also varies but FormBook, Agent Tesla, NanoCore and Netwire are among the most common.
Symantec has multiple 0-day protections in place for GuLoader. Here are the current top 10:
File-based
Exp.CVE-2017-11882!g7
Packed.NSISPacker!g14
Scr.Malcode!gen19
Trojan Horse
Trojan.Gen.2
Trojan.Gen.MBT
Machine Learning-based
Heur.AdvML.A!300
Heur.AdvML.A!400
Heur.AdvML.B!200
Heur.AdvML.C
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers.
To learn more about Symantec's Endpoint Security Service, click here.
To learn more about how Symantec Endpoint Protection uses advanced machine learning, click here.
To learn more about Symantec's Cloud Email Security Service, click here. |
Protection Highlight: Hive ransomware hits 64-bit 0-day roadblock
Hive is a particularly persistent ransomware first seen in June 2021 and still one of the most prevalent variants as of November 2022, believed to have affected hundreds of organizations to date. According to a recently published joint Cybersecurity Advisory (CSA), more than 1,300 victims globally and over $100 million in ransom money.
Known to enter a victim's network through by various means including Vulnerable RDP servers, compromised VPN credentials and Phishing emails, once inside the network it deploys remote access software such as AnyDesk or ScreenConnect to maintain persistence within the network, leveraging legitimate remote command execution tools like PSExec, and combinations of LOLBins such as WMI and BITSAdmin to spread laterally across the network.
While originally written in Goland, Hive has since been ported to the Rust programming language which allows the attackers to target Linux-based operating systems, including VMWare ESXi virtual machines and servers. As with many other ransomware variants, Hive employs the double extortion tactic meaning it not only encrypts the victim's files locally, it also exfiltrates them to a remote location and threatens to make them public if the ransom is not paid.
Recently however, new Hive variants have found themselves up against Symantec's 64-bit machine-learning technology, and effectively being stopped dead in their tracks. To be clear, this is proactive defense, blocking previously unseen variants without requiring the addition of reactive signatures or updated protections.
Symantec has 0-day protection in place, identified by the following:
Machine Learning-based
Heur.AdvML.B
To learn more about how Symantec Endpoint Protection uses advanced machine learning, click here. |
Protection Highlight: How Adaptive Protection Empowers Security Teams to Thwart Evolving Threats
As a Chief Information Security Officer (CISO) or a member of the Security Operations Center (SOC) in a company, the threat landscape constantly presents evolving challenges, with ransomware attacks and "Living off the Land" tactics being particularly pervasive and damaging. These threats often leverage legitimate tools and processes within an organization, making them difficult to detect and mitigate without disrupting essential business operations.
Symantec's Adaptive Protection feature offers a crucial solution to this dilemma. By allowing customers to monitor and block generic application behaviors from occurring in their company, while allowing legitimate use cases through automatic whitelisting. It empowers organizations to proactively defend against ransomware and other malicious activities while minimizing disruptions to daily operations.
Analyzing recent security incidents reveals a concerning trend where attackers like AlphaV, Lockbit and Play Ransomware leverage a wide array of generic application behaviors to carry out their malicious activities. Exploiting tools like Screenconnect, PsExec, and VssAdmin for unauthorized system access, lateral movement within networks, and system manipulation. Similarly, they utilize 7Zip for concealing malicious payloads and AnyDesk for unauthorized system access, enabling data exfiltration and malware deployment. Additionally, attackers leverage PowerShell, PsExec, and WinRAR for executing commands, deploying malware, escalating privileges, and obfuscating payloads or exfiltrating data.
To illustrate how a customer utilizes our Adaptive Protection feature, consider this example from September 2023. A customer already had 126 behaviors set to Deny and aimed to bolster their security posture further. The steps they took were as follows:
Utilized the Adaptive Protection’s heatmap to assess behavior insight within their environment.
Identified a set of behaviors that were in monitoring mode and were not used in the company. This was done by identifying the dark blue (zero prevalence) entries on the heatmap page.
Used the threat overlay filter to spot any risky behaviors that were not dark blue but were considered for “deny”.
Examined system-generated exceptions to whitelist normal business operational behaviors.
Updated the Adaptive Policy by adding new behaviors to the "Deny" mode.
While not all adaptive-based signatures were toggled to deny, this customer's selection gave them an extra layer of protection against a well-known Ransomware group known as Lockbit, which keeps on resurfacing despite takedown efforts. Below is a breakdown of the Living off the Land portion of the group's attack as of January 19, 2024, that Symantec's Adaptive Protection monitors:
Untrusted process updates registry to setup an autorun
Untrusted process updates task scheduler
Disable security settings through Netsh
Dump OS credentials
Use Netscan to gather information on the environment
Use Psexec to execute code
Use rclone to move data out of the network
Powershell creates an executable file
Bcdedit to disable recovery
Powershell calls VSSAdmin to delete backups
Powershell calls WBadmin to delete backups
Fsutil erases the traces of ransomware executable
Powershell downloads a payload
This attack relies on many system tools for a variety of tasks, from code execution to cleaning up the traces of the malware's existence. Someone may look at this list and think they could turn on Deny for all of these behaviors in their environment—great! However, every customer’s needs are different, and Adaptive Protection provides the flexibility to ensure maximum security with minimal interruptions.
In conclusion, the use of Adaptive Protection among our customers demonstrates its effectiveness and widespread adoption.
469 behaviors available to set to Deny mode
On average, a customer has 338 behaviors set to Deny.
Over 1.5 million machines are currently protected by this feature.
This results in approximately 8,000 behaviors being blocked each day.
While the counts of blocked behaviors might seem modest, it's crucial to recognize that these behaviors are associated with standard or normal tools, not malware. Such activities might go undetected by traditional security technologies, yet they have the potential to cause significant damage if not appropriately managed. This highlights the importance of Adaptive Protection in identifying and mitigating threats that leverage everyday tools, ensuring a more secure environment.
To learn more about Symantec's Endpoint Security Service, click here.
To learn more about Symantec's Adaptive Protection feature, click here. |
Protection Highlight: HTML-based Attachments, A Growing Trend in Phishing
In the ever-evolving landscape of cyber threats, few adversaries have stood the test of time like phishing. Almost as old as the internet itself, phishing attacks continue to plague organizations and individuals worldwide. Unlike complex cyber exploits that exploit technical vulnerabilities, phishing preys on the human element, capitalizing on psychological manipulation to achieve its objectives.
While the internet has seen a proliferation of new communication channels and technologies, email remains the primary vector for phishing attacks, although malicious SMS have been on the rise. When it comes to emails, HTML-based phishing, encompassing HTML, SHTML, and XHTML, persists as one of the most prevalent techniques. Its enduring popularity is primarily due to its ease of creation and customization. Cybercriminals, often with minimal technical expertise, can swiftly create convincing replicas of legitimate websites using readily available templates and online tutorials.
In most cases if not all, the use of persuasive language, fake security alerts, and urgency-inducing messages enhances the illusion of legitimacy, pre-filled login field with recipients email address, makes it challenging for users to discern between genuine and fake websites. Moreover, attackers employ various obfuscation techniques to conceal malicious code within HTML-based phishing pages. Encoding, encryption, and dynamic content generation are utilized to evade traditional detection mechanisms and bypass security controls.
Phishing campaigns come in various sizes and target different victim profiles. These victims may be either specifically targeted or part of a broader non-targeted group, with certain commonalities such as geographic regions or industry sectors. Although phishing attacks might seem less impactful compared to high-profile cyber threats like ransomware or cyberespionage, they serve as the crucial initial steps for many cyberattacks. They lead to a myriad of detrimental outcomes, including data theft, financial losses, identity theft, ransomware infections, credential stuffing attacks, business email compromise, network compromise, and reputational damage.
Symantec's Cloud Email Security Service and Messaging Gateway stand as a barrier against this type of phishing and phishing in general, thanks to a comprehensive suite of advanced technologies meticulously designed to safeguard organizations against such threats.
Anti-Spam filtering system: a predictive filtering systems focused on different attributes of attachment files and other email features are deployed in timely intervals to catch changes in the rapidly changing email threat landscape.
Emulators and Heuristic Signatures: Our solution utilizes sophisticated emulators and heuristic signatures to detect malicious HTMLs embedded within emails. By simulating the execution of code and analyzing behavioral patterns, we can identify and neutralize potential threats before they can harm an organization.
URL reputation via WebPulse technology: A cloud-based infrastructure specifically designed to harness the power of user-driven behavior and to translate user input into global web intelligence and web threat intelligence.
Data protection service: HTML, SHTML, and XHTML – we recommend blocking those outright with a data protection rule if there's not a legit business reason for these mime types.
We understand that the threat landscape is constantly evolving, with cybercriminals devising new tactics to evade detection. That's why we maintain a vigilant stance, continuously monitoring the threatscape to identify emerging phishing campaigns with unique twists in their methodologies.
To learn more about Symantec's Cloud Email Security Service, click here.
To learn more about Symantec's Messaging Gateway, click here.
To learn more about Symantec's cloud-based Web Security Service (WebPulse), click here. |
Protection Highlight: Hunters International Ransomware
Over the past few months, several new ransomware actors have been observed targeting businesses of all sizes in various countries around the world. These include groups like Hunters International, Meow, DragonForce, Werewolves, Malekteam, etc. All of these actors are being tracked by Symantec, whether they engage in single extortion through exfiltration or encryption, or use double-extortion involving both exfiltration and encryption.
In this protection highlight, we will cover Hunters International, which made the news in late 2023 after claiming multiple victims, and moving into 2024 has claimed even more. This actor is using a ransomware that appears to be closely based on the infamous Hive ransomware code, of which the original group was disrupted by law enforcement in early 2023.
As of today, the Hunters International group's full tactics, techniques, and procedures remain incomplete analysis-wise. However, they are known to move laterally within the victim's infrastructure and upon finding what they consider to be worthy targets, they exfiltrate sensitive data and encrypt files - a classic modus operandi of most double-extortion ransomware actors.
If the ransomware binary (collected in January) is successfully triggered on compromised machines, similar to previous samples, it will attempt to kill processes and services. Next, it executes commands to delete backups and disable recovery mechanisms. Then, it iterates through local and mapped drives, as well as shared drives found on the local network through the NetServerEnum and NetShareEnum APIs, encrypting files that are discovered. It adds the extension .lock to each encrypted file and drops a ransom message named "Contact Us.txt" in the same directory. Although the goal is to encrypt files, it will skip the following:
Files in folders with these substrings in their names: $Recycle bin, $windows.~bt, $windows.~ws, all users, appdata, boot, config.msi, default, google, intel, mozilla, msocache, perflogs, system volume information, tor browser, internet explorer, windows, windows.old, windows nt
Files named: autorun.inf, bootfont.bin, boot.ini, bootsect.bak, desktop.ini, iconcache.db, ntldr, ntuser.dat, ntuser.dat.log, ntuser.ini.log, thumbs.db
Files with the following extensions: 386, adv, ani, bat, bin, cab, cmd, com, cpl, cur, deskthemepack, diagcab, diagcfg, diagpkg, dll, drv, exe, hlp, hta, icl, icns, ico, ics, idx, key, ldf, lnk, lock, mod, mpa, msc, msi, msp, msstyles, msu, nls, nomedia, ocx, pdb, prf, ps1, rom, rtp, scr, shs, spl, sys, theme, themepack, wpx
A screenshot of the ransom note left on the victim's machine.
As mentioned above, Symantec constantly tracks new ransomware actors by monitoring new ransom/leak websites and other data from both external and internal sources. While the malware and tools used are not always readily available or easily found, we continue to actively seek samples and identify all TTPs. It's a constant cat-and-mouse game where staying informed and obtaining indicators of compromise (IOCs) are critical components of effective threat detection and mitigation.
Symantec protects you from this threat, identified by the following:
File-based
Ransom.Hunters!g1
Trojan.Gen.MBT
Trojan Horse
Behavior-based
SONAR.SuspLaunch!gen4
SONAR.SuspLaunch!g18
SONAR.SuspLaunch!g195
SONAR.SuspLaunch!g253
SONAR.RansomPlay!gen1
SONAR.RansomGen!gen3
SONAR.SuspLaunch!g193
Machine Learning-based
Heur.AdvML.B!100
Policy-based
Symantec DCS Hardening policy for Windows provides 0-day protection against Hunter's Hive ransomware. The default sandboxing controls prevent webshells and malware tools from getting installed and prevent arbitrary system command execution from privileged applications.
EDR-based
Symantec EDR is capable of monitoring and flagging this threat actor's tactics, techniques and procedures.
Symantec added malware-specific threat hunting queries which can be fired on the customer iCDM console. For more information, refer to this link: https://github.com/Symantec/threathunters/tree/main/Ransomwares/HunterInternational/
Click here to learn more about Symantec's Endpoint Security Service.
Click here to learn how Symantec behavioral security technologies provide protection against zero-day attacks.
Click here to learn more about how Symantec Endpoint Protection uses advanced machine learning.
Click here to learn more about Symantec's Data Center Security.
Click here to learn about Symantec Endpoint Detection & Response.
** Symantec Endpoint Detection and Response (EDR) uses machine learning and behavioral analytics to detect and expose suspicious network activities. EDR alerts about potentially harmful activity, prioritizes incidents for quick triage, and allows incident responders to navigate device activity records for purposes of forensic analysis of potential attacks. |
Protection Highlight: IcedID malicious campaigns across July and August
IcedID (also known as Bokbot) is a well known modular banking malware that initially emerged in the threat landscape around 2017. Over time it evolved and nowadays is more often used as a loader for additional malicious modules and payloads, including ransomware. IcedID is known to be widely spread by spear-phishing campaigns. In the past, it also was often delivered as a secondary payload during infections with other malware variants such as Emotet.
Symantec has observed some uptick in IcedID related activities across July and August as a result of multiple malspam campaigns that are still on-going. The most recent runs from August exhibit a multi-staged attack chain that involves PDF documents with filenames such as "Doc_Scan_08_18" or "Document_08_22" delivered by malspam. The malicious PDFs purport to be secure invoices requiring the victims to download the intended document by clicking the "download" or "open" button as seen on the figure below:
Upon clicking the link, the victim gets redirected to a URL containing malicious JavaScript, that in turn once executed delivers the IcedID payload in the form of a .dll binary onto the infected machines.
The below chart shows the daily breakdown of the observed IcedID activity over the past two months.
Symantec has 0-day protection in place against this threat, identified by the following:
Behavior-based
SONAR.IcedID!g4
SONAR.IcedID!g5
SONAR.SuspLaunch!g235
SONAR.TCP!gen1
EDR-based
Symantec EDR is capable of monitoring and will flag this threat actor's tactics, techniques and procedures
Symantec added malware specific threat hunting queries which can be fired on customer iCDM console. For more information, you can refer to this link: https://github.com/Symantec/threathunters/tree/main/Trojan/IcedID
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers.
File-based
Scr.DLHeur!gen1
Scr.IcedID!gen1
Scr.IcedID!gen3
Scr.Malcode!gdn28
Scr.Malcode!gen46
Scr.Malpdf!gen2
Trojan.IcedID
Trojan.IcedID!g16
Trojan.IcedID!g17
Trojan.IcedID!g18
Trojan.IcedID!gm
Trojan.Gen.MBT
Trojan Horse
WS.Malware.1
Machine Learning-based
Heur.AdvML.A!300
Heur.AdvML.A!400
Heur.AdvML.A!500
Heur.AdvML.B!100
Heur.AdvML.B!200
Heur.AdvML.C
Network-based
System Infected: Trojan.Backdoor Activity 592
System Infected: Trojan.Backdoor Activity 634
System Infected: Trojan.Backdoor Activity 764
System Infected: Trojan.Backdoor Activity 765
Web Attack: Webpulse Bad Reputation Domain Request
Web-based
Observed domains/IPs are covered under security categories in all WebPulse enabled products.
To learn more about Symantec's Cloud Email Security Service, click here.
To learn more about Symantec's cloud-based Web Security Service (WebPulse), click here.
To learn how Symantec behavioral security technologies provide protection against zero-day attacks, click here.
To learn more about Symantec's Endpoint Security Service, click here.
To learn more about how Symantec Endpoint Protection uses advanced machine learning, click here.
To learn more about managing Symantec's Endpoint Security Intrusion Prevention System (IPS), click here. |
Protection Highlight: Inc Ransomware
Symantec recently observed an emerging ransomware group that goes by the name "Inc. Ransom." While their attack chain is still partially unknown, there have been indications that they leverage the Remote Desktop Protocol (RDP) to connect to target machines, and subsequently use legitimate tools to view, archive and extract documents.
Upon successful encryption, Inc ransomware will append a .inc extension to encrypted files. It will also drop ransom notes in both HTML and TXT format (INC-README.txt and INC-README.html) in various locations in the compromised machine's directory, threatening to publish the victim's stolen data if they don't contact the attackers within 72 hours.
Symantec has 0-day protection in place, identified by the following:
File-based
Ransom.Inc
Trojan.Gen.MBT
Behavior-based
SONAR.Ransomware!g1
SONAR.Ransomware!g7
Machine Learning-based
Heur.AdvML.B
Policy-based
Symantec Data Center Security (DCS) default hardening policies provide 0-day protection against Inc ransomware
EDR-based
Symantec EDR is capable of monitoring and flagging this threat actor's tactics, techniques and procedures.
Symantec added malware-specific threat hunting queries which can be fired on the customer iCDM console. For more information, refer to this link: https://github.com/Symantec/threathunters/tree/main/Ransomwares/Inc
Click here to learn more about Symantec's Endpoint Security Service.
Click here to learn how Symantec behavioral security technologies provide protection against zero-day attacks.
Click here to learn more about how Symantec Endpoint Protection uses advanced machine learning.
Click here to learn more about Symantec's Data Center Security.
Click here to learn about Symantec Endpoint Detection & Response. |
Protection Highlight: IPFS Phishing On The Rise
IPFS, short for InterPlanetary File System, is a peer-to-peer distributed file system allowing people around the world to exchange files. Instead of using file paths like location-based centralized systems, IPFS uses unique content identifiers (CID) generated based on the file’s hash value, with the file residing on the computer of the user who shared it with the IPFS. Files shared in this way are downloaded directly from that computer, hence the term "distributed". IPFS effectively replaces the familiar HTTP and HTTPS protocols (Hypertext Transfer Protocol / Secure) as a different way to access the World Wide Web.
IPFS phishing attacks are similar to phishing attacks conducted through a centralized network where attackers use social engineering techniques, email and messaging platforms, and fake websites to impersonate legitimate brands in order to steal your credentials and/or gain access to your network. Using a unique IPFS address, content published on the IPFS network becomes accessible to anyone, who can then republish it on their own node making it easy for attackers to create permanent and untraceable phishing sites that remain active even after the original source has been taken down.
To facilitate their attacks, malicious actors may employ one of the following methods:
* Malicious URL: A phishing text message, email, direct message, pop-up or other means of tricking a user into clicking a link leading to a malicious IPFS gateway.
* DNS (Domain Name System) spoofing: Creating a fake DNS server that redirects to a malicious IPFS gateway hosting a fake website.
* SSL (Secure Sockets Layer) certificate: A fake SSL certificate which convinces the user they are visiting a legitimate site.
Here's an example of a phishing page that looks like an excel file, prompting the user to enter their password in order to access it.
The IPFS offers several benefits over centralized file sharing, but over the years it has also become an attractive platform for cybercriminals. Symantec observes tens of thousands of IPFS phishing emails every month.
Separately, over the past 30 days Symantec's Web Security Service has flagged thousands of CID subdomains including 11,789 unique hosts, 19,071 unique full URLs, and 55,164 CID paths.
Symantec has 0-day protection in place against IPFS attacks, identified by the following:
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers.
Web-based
Observed domains/IPs are covered under security categories. Newly observed URLs are detected and categorized using known IPFS URL patterns.
To learn more about Symantec's Cloud Email Security Service, click here.
To learn more about Symantec's cloud-based Web Security Service (WebPulse), click here. |
Protection Highlight: IPS Audit Signatures - digital watchdogs against Discord and Telegram abuse
Discord and Telegram are popular messaging and file-sharing applications that have gained significant traction in recent years. Discord is primarily geared towards gaming communities and offers a variety of features catering to this demographic, including voice and text chat channels, file sharing, and customizability. Telegram on the other hand, is a more general-purpose messaging app that provides end-to-end encryption for messages, allowing users to communicate securely.
Due in large part to their widespread popularity, both Discord and Telegram have become regular targets of cybercriminals for malicious purposes. Several factors contribute to this trend:
Large user base: Both Discord and Telegram boast millions of active users, making them attractive targets for cybercriminals seeking to reach a broad audience.
File sharing capabilities: Both platforms allow users to easily share files, including executable files, which can be used to distribute malware.
Ease of use: Both Discord and Telegram are relatively easy to use, making them accessible to users with varying levels of technical expertise including those with malicious intent.
Lack of moderation: While both platforms have implemented some moderation measures, the sheer volume of user-generated content likely makes it difficult for them to effectively identify and remove malicious content.
As a result of these factors, Discord and Telegram have become increasingly popular channels for malicious file hosting and stolen information exfiltration. Cybercriminals use these platforms to share malware, distribute stolen data, and engage in other illicit activities. Our readers can view previous protection bulletins related to threats potentially taking advantage of Discord and/or Telegram here.
In response to this increased activity - in addition to strong AV protections (behavioral, heuristic, and machine learning) - Symantec has released several IPS Audit Signatures to provide an extra layer of security, giving customers an edge in monitoring unvetted and potentially malicious Discord and Telegram network traffic known to be used by malicious actors.
Symantec IPS Audit Signatures reported the following activity over the past 2 month period
SEP's Audit Signatures are intended to raise awareness of potentially unwanted traffic in the network. By default, they do not block. Administrators reviewing the logs of IPS events in their network can note these Audit events and decide whether or not to configure the corresponding Audit Signatures to block the traffic.
Symantec protects you from these threats, identified by the following:
Network-based
Audit: PowerShell Process Accessing discordapp
Audit: System Process Accessing discordapp
Audit: Untrusted Telegram API Connection
To learn more about managing Symantec's Endpoint Security Intrusion Prevention System (IPS), click here. |
Protection Highlight: IPS blocking Tech Support scams daily
It feels like technical support scams have been around forever and you'd be forgiven for thinking people would no longer fall victim to them. Sadly, that's not the case. Our network traffic monitoring systems show these scams are still very much alive and kicking. Symantec continues to observe new fake tech support websites and malicious traffic redirections to scammer controlled domains on a daily basis.
Technical support scams affect both consumers and enterprises, but they can be a lot more impactful to organizations and businesses due to the potential for theft of sensitive company-related information. While financial gain is usually the primary goal, impersonating a technical support representative and convincing the employee to allow remote access to their computer or expose sensitive information can lead to data breaches and serious financial loss. Successful tech support scams also contribute to the collection of stolen credentials which are then sold on black markets resulting in additional opportunities for fraud to be committed.
Symantec has 0-day protection in place against this malicious network traffic, identified by the following:
Network-based
Web Attack: Fake Scan Webpage*
Malicious Site: Malicious Domain Request*
Web Attack: Fake Tech Support Domains*
Web Attack: Fake Tech Support Website*
* This indicates multiple similarly named detections exist such as Web Attack: Fake Tech Support Website 295, Web Attack: Fake Tech Support Website 374, etc.
Web-based
Observed domains/IPs are covered under security categories.
To learn more about managing Symantec's Endpoint Security Intrusion Prevention System (IPS), click here.
To learn more about Symantec Web Protection, click here.
To learn more about Symantec's cloud-based Web Security Service (WebPulse), click here. |
Protection Highlight: IPS Blocks SMB Attacks
The Server Message Block (SMB) protocol is a client-server communication protocol used primarily by the Windows operating system but also on Linux and macOS computers to share files, printers, and other resources hosted on the same network. Developed by IBM back in the early 1980s and immediately recognizable to network engineers by its use of TCP ports 139 and 445, to this day SMB continues to be one of the most common methods of sharing files in the workplace. But despite the protocol having been updated multiple times over the years to meet evolving network requirements, many devices are still running on the older, less secure versions, inevitably making it a prime target for cyber criminals.
SMB attacks are a type of cyber attack that target vulnerabilities in the SMB protocol in order to gain unauthorized access to a network and help facilitate lateral movement within the network (lateral movement is a tactic used by attackers to move beyond the initial point of entry to gain access to additional network resources). Two of the more infamous SMB attacks include the 2017 "EternalBlue" exploit (CVE-2017-0144) which was leveraged to great effect by the equally infamous "WannaCry" ransomware, and the more recent SolarWinds attack which also exploited vulnerabilities in the SMB protocol.
Some common SMB attacks include:
* SMB brute-force attacks: This involves repeatedly trying different username and password combinations to gain access to a target system. This can be done manually or with the help of automated tools.
* SMB relay attacks: A type of man-in-the-middle (MITM) attack, the attacker intercepts SMB traffic between two machines and then relays the traffic to a third machine controlled by the attacker. This allows the attacker to gain access to the target system without the need for valid credentials.
* SMB worm attacks: These attacks use malware that spreads through a network by exploiting vulnerabilities in the SMB protocol. Once the malware infects a system, it can spread to other systems on the network.
* SMB denial-of-service attacks: This type of attack involves overwhelming a target system with a flood of SMB requests, causing the system to crash or become unavailable.
SMB vulnerabilities have long been a favorite of malicious actors to conduct attacks on corporate systems spanning the globe and Symantec's IPS technology blocks millions of SMB compromise attempts on a daily basis.
Symantec has 0-day protection in place against a wide range of SMB attacks, identified by the following:
Network-based blocking signatures
Attack: Bluwimps SMB Activity
Attack: Fake SMB Server Response
Attack: SMB Arbitrary Service Create Request 2
Attack: SMB Double Pulsar Ping
Attack: SMB Double Pulsar Response
Attack: SMB PE File Drop Startup Directory
OS Attack: Microsoft SMB MS17-010 Disclosure Attempt
OS Attack: Microsoft SMB MS17-010 Shellcode Attempt
OS Attack: Microsoft Windows SMB RCE CVE-2017-0144
OS Attack: Microsoft Windows SMB Remote Code Execution CVE-2017-0143*
OS Attack: Microsoft Windows SMB Remote Code Execution CVE-2017-0144*
OS Attack: MS SMB2 Validate Provider Callback CVE-2009-3103
OS Attack: SMB EFS NTLM Relay Attempt
OS Attack: SMB Validate Provider Callback CVE-2009-3103
OS Attack: Windows SMBv3 CVE-2020-1206
System Infected: Bad Reputation File SMB Request
Network-based audit signatures**
Audit: Bad Reputation File SMB Request
Audit: Microsoft Compressed SMB Packet
Audit: SMB Admin Share Connect Request
Audit: SMB Bruteforce Attempt
Audit: SMB Exchange Server WebShell Access Attempt
Audit: SMB Request From External Host
Audit: SMB Suspicious DLL Create Attempt
Audit: SMB Suspicious Folder File Creation
Audit: SMB Unimplemented Trans2 Subcommand
Audit: SMB Windows Print Spooler RpcAddPrinterDriverEx Attempt
Audit: SMBv1 NTLM Authentication Attempt
Audit: SMBv1 Traffic Request
Audit: SMBv2 NTLM Authentication Attempt
Audit: Suspicious SMB Client Activity
Audit: Suspicious SMB Client Request*
Audit: Suspicious SMB Server Response
* This indicates multiple similarly named detections exist such as Audit: Suspicious SMB Client Request2, Audit: Suspicious SMB Client Request3, etc.
** SEP's Audit Signatures are intended to raise awareness of potentially unwanted traffic in the network. By default, they do not block. Administrators reviewing the logs of IPS events in their network can note these Audit events and decide whether or not to configure the corresponding Audit Signatures to block the traffic.
To learn more about managing Symantec's Endpoint Security Intrusion Prevention System (IPS), click here. |
Protection Highlight: IPS Chrome Browser Extension
Through our regular "What did IPS do for you last week" bulletin series we attempt to convey that IPS (Intrusion Prevention System) is a critical protection layer in the fight against malware, including both malicious files and malicious network traffic. IPS blocks 70% of all malware - far more than traditional file-based technologies. Within that 70%, close to 95% are what we refer to as "pre-infection" blocks, meaning these attacks were prevented from entering the Enterprise environment at all, as opposed to something that managed to get past at least one layer of security but was prevented by IPS from spreading further inside the network or communicating externally.
This outbound blocking is also an important protective mechanism as it stops the download of additional malware (possibly the actual payload), communications with the attacker's command-and-control server (i.e. phoning home) and the exfiltration of company data. If your security setup doesn't include Intrusion Prevention you are potentially exposing your organization to a significant loss of threat protection.
A very useful IPS-related feature some people may not be aware of is the Chrome browser extension. Symantec Endpoint Protection 14.3 RU2 and later includes a Google Chrome extension to protect client computers from accessing malicious websites. The Chrome browser extension monitors inbound and outbound HTTP and HTTPS traffic to the web browser and blocks the traffic if the client determines that the URL is malicious. It does this through a combination of IPS signatures and URL reputation lookups (via our WebPulse and Insight services) to identify threats from domains and URLs that host malicious content such as malware, fraud, phishing, and spam.
Each spike represents new weekly campaigns involving malicious domains and associated web traffic.
Symantec has protection in place against this malicious network traffic, identified by the following:
If the Chrome web browser detects that a URL is malicious, the client redirects users to the following default landing page:
To learn more about the Symantec Endpoint Protection Chrome browser extension, click here. |
Protection Highlight: IPS for CMS Vulns
CMS platforms such as WordPress, Joomla, Drupal, and others are widely used to create and manage websites. Their popularity means that a significant number of websites are built on these platforms, making them attractive targets for attackers. These platforms are complex software systems with various components, themes, plugins, and modules, often created by different developers. This complexity can introduce vulnerabilities that can be exploited by threat actors. With numerous third-party themes and plugins available, it becomes challenging to ensure the security of every component.
Unfortunately many website owners don't keep their CMS and its components up-to-date, which can leave vulnerabilities unpatched making it easier for attackers to exploit known security flaws. Usually CMS vulnerabilities are well-documented and attackers actively search for websites running outdated versions. Most actors would leverage automated tools to exploit these, making it easier to launch widespread attacks. They can scan the Internet for websites running specific CMS versions, identify vulnerabilities, and launch automated attacks at scale.
Symantec's Intrusion Prevention System (IPS) has been proactively blocking threats which attempt to exploit these CMS vulnerabilities.
Servers are particularly vulnerable. In our "What did IPS do to protect Servers last month?" bulletin for May we reported 1.4M attempts to scan for CMS Vulnerabilities blocked on 12.2K servers. The attackers never let up.
Over the years, CMS vulnerabilities have been associated with various threats and malicious activities. Some of the common threats that have emerged include:
Unauthorized Access: Exploiting CMS vulnerabilities can provide attackers with unauthorized access to a website's backend, allowing them to gain control over the content management system. This access can be used to manipulate website content, steal data, or carry out other malicious activities.
Defacement: CMS vulnerabilities can be exploited to deface websites by injecting unauthorized content or modifying existing content. Attackers may replace the website's original content with malicious or offensive material, causing reputational damage to the website owner or organization.
Data Breaches: CMS vulnerabilities can lead to data breaches, where attackers gain access to sensitive information stored on the website. This can include user credentials, personal information, financial data, or proprietary business information. Stolen data can be exploited for identity theft, financial fraud, or sold on the black market.
Malware Distribution: Exploiting CMS vulnerabilities can allow attackers to inject malicious code into a website. This code can be used to distribute malware to unsuspecting visitors, compromising their devices and potentially spreading the infection to other systems.
Phishing Attacks: CMS vulnerabilities can be exploited to set up phishing campaigns. Attackers can create fake login pages or forms that appear legitimate but are designed to steal user credentials or sensitive information. Phishing attacks can lead to identity theft, financial loss, or unauthorized access to other accounts.
SEO Spamming: CMS vulnerabilities can be leveraged to inject hidden links or keywords into a website's content. This is often done for SEO (search engine optimization) spamming purposes, where attackers manipulate search engine rankings to promote their own websites or products.
Distributed Denial-of-Service (DDoS) Attacks: CMS vulnerabilities can be exploited to launch DDoS attacks against websites. By exploiting weaknesses in the CMS or associated plugins, attackers can overload the website's servers with a massive volume of requests, rendering the website inaccessible to legitimate users.
Symantec has multiple 0-day protections in place against these attacks, including but not limited to the following:
Network-based
Web Attack: WordPress Plugin XSS Attempt
Web Attack: WordPress XMLRPC Malicious Pingback Request
Web Attack: Wordpress Arbitrary File Download 4
Web Attack: Sourcecodester System CVE-2020-29227
Web Attack: Joomla Component Local File Inclusion
Attack: Web CMS Multiple Sql Injection
Web Attack: Drupal Core RCE CVE-2018-7602
Web Attack: WordPress Plugin Path Traversal Attempt
Web Attack: Wordpress Arbitrary File Download CVE-2003-1599
Attack: Wordpress Duplicator Plugin Unauthenticated Arbitrary File Download
To learn more about managing Symantec's Endpoint Security Intrusion Prevention System (IPS), click here. |
Protection Highlight: LockBit - Prolific, Persistent, Preventable
LockBit is a ransomware-as-a-service (RaaS) operating as early as September 2019 by malicious actors Symantec tracks as Syrphid. Attacking organizations of varying sizes including but not limited to financial services, professional services, food and agriculture, legal, education, energy, emergency services, healthcare, manufacturing and transportation, it is believed to be responsible for up to 1700 attacks to date including 1 in 6 ransomware attacks on the U.S. government during 2022. LockBit employs the RaaS model where new affiliates are recruited to conduct attacks using LockBit ransomware tools and infrastructure. This diverse range of affiliates results in LockBit ransomware attacks varying widely in it's observed tactics, techniques, and procedures (TTPs) making it all the more difficult for organizations to defend against.
LockBit has continued to evolve over the years. Originally calling itself “ABCD ransomware" upon its initial release in 2019, the affiliate program launched in January 2020, then in September of 2020 a leak site (where they list compromised organizations who refused to pay the ransom) was introduced, and LockBit also adopted the double extortion model. In 2021 LockBit 2.0, leveraging an older vulnerability identified as CVE-2018-13379 was seen targeting Australian organizations. Toward the end of 2021 a Linux variant was released focusing on VMware ESXi virtual machines. June 2022 saw the 3.0 variant added, referred to as "LockBit Black" due to code similarities to BlackMatter & Darkside ransomware variants. LockBit 3.0 can delete a number of pre-defined services and terminate certain processes.
In July of the same year attackers were observed using a tool called Terminator to attempt to disable security software. These types of attacks, known as Bring Your Own Vulnerable Driver (BYOVD) attacks, involve legitimate drivers signed with valid certificates which are capable of running with kernel privileges, being dropped onto victim devices to disable security solutions and take over the system.
Symantec customers should note that any commands or tools attempting to stop SEP will only work if the attacker has administrative credentials and the SEP administrator has disabled anti-tamper protection.
As recently as April this year a new LockBit variant targeting the macOS platform was discovered, with the samples seemingly based on LockBit’s Linux encryptor and only compiled for macOS. The latest intel from security researchers is that the LockBit operation appears to have started experimenting with new versions of its payload, capable of attacking multiple architectures, including Apple M1, ARM v6, ARM v7, FreeBSD, and others.
A dangerous and constantly evolving threat such as LockBit demands an equally aggressive response.
Symantec has a comprehensive range of protections in place against LockBit's many variants and their individual components.
File-based
Downloader
OSX.Ransom.Lockbit
Packed.Generic.686
Ransom.Lockbit
Ransom.Lockbit!g1
Ransom.Lockbit!g2
Ransom.Lockbit!g6
Ransom.Lockbit!g7
Ransom.Lockbit!gen3
Ransom.Lockbit!gen4
Ransom.Lockbit!gen5
Ransom.Lockbit!gm1
Scr.Malscript!gen1
Trojan.Gen.2
Trojan.Gen.MBT
Trojan Horse
WS.Malware.1
Behavior-based
Ransom.Blackmater!gm1
SONAR.Cryptlocker!g42
SONAR.ProcHijack!gen5
SONAR.ProcHijack!g45
SONAR.Ransomware!g2
SONAR.Ransomware!g7
SONAR.RansomLckbit!g1
SONAR.RansomLckbit!g2
SONAR.RansomLckbit!g3
SONAR.RansomLckbit!g4
SONAR.RansomLckbit!g5
SONAR.RansomNokibi!g1
SONAR.SuspBeh!gen82
SONAR.SuspBeh!gen742
SONAR.SuspLaunch!gen4
SONAR.SuspLaunch!gen18
SONAR.SuspLaunch!g189
SONAR.SuspLaunch!g190
SONAR.SuspLaunch!g193
SONAR.SuspLaunch!g195
SONAR.SuspLaunch!g253
SONAR.SuspReg!gen28
SONAR.UACBypass!gen30
Machine Learning-based
Heur.AdvML.B
Heur.AdvML.C
Heur.AdvML.B!100
Heur.AdvML.B!200
Network-based
Attack: Ransom.Lockbit Activity
Attack: Ransom.Lockbit Activity 2
Attack: Ransom.Lockbit Activity 3
Web Attack: Webpulse Bad Reputation Domain Request
Web Attack: Fortinet FortiOS Directory Traversal CVE-2018-13379
Attack: Lockbit Ransomware Binary Copy GPO Config
Attack: Lockbit Ransomware Enable Share GPO Config
Attack: Lockbit Ransomware Security Services Taskkill GPO
Attack: Lockbit Ransomware Services Disable GPO Config
Policy-based
DCS default hardening policy provides zero day protection against LockBit ransomware.
Suspicious Process Execution: Prevention policies prevent this malware from being dropped or executed on the system.
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection.
Web-based
Observed domains/IPs are covered under security categories in all WebPulse enabled products.
To learn more about Symantec's Endpoint Security Service, click here.
To learn how Symantec behavioral security technologies provide protection against zero-day attacks, click here.
To learn more about how Symantec Endpoint Protection uses advanced machine learning, click here.
To learn more about managing Symantec's Endpoint Security Intrusion Prevention System (IPS), click here.
To learn more about Symantec's Data Center Security, click here.
To learn more about Symantec's Cloud Email Security Service, click here.
To learn more about Symantec's cloud-based Web Security Service (WebPulse), click here. |
Protection Highlight: Lokibot still a danger
Dating back to around 2015 and still very active today, the Lokibot infostealer spreads via a continuous stream of spam campaigns, often leveraging malicious PDFs, RTFs and Office documents attached to the email as the infection vector and using quotes, shipping, banking, SWIFT, invoice and payment related social engineering themes, Lokibot attempts to steal credentials from hundreds of applications including browsers, FTP clients, Email clients, SSH clients, cryptocurrency wallets and password management software and may use one of several different packing methods for obfuscation, but ultimately has to unpack itself into memory prior to execution of the main payload which is often where security products will block it.
Symantec recently observed new campaigns involving emails with a variety of subjects including quotes, confirmation of payment, new contracts, purchase orders, etc.
As usual, the emails contain an attachment being either an archive file (rar, gz, etc.) or a Microsoft Office file. In all cases the attachments contain an executable file which hosts the actual Lokibot payload. The following example shows an email containing a .gz file - an archive compressed using the gzip compression technology.
Symantec has multiple 0-day protections in place, identified by the following:
File-based
Infostealer.Lokibot
Infostealer.Lokibot!12
Infostealer.Lokibot!16
Infostealer.Lokibot!22
Infostealer.Lokibot!34
Infostealer.Lokibot!43
Infostealer.Lokibot!gm
Scr.Malcode!gen59
Scr.Malcode!gdn32
Scr.Malcode!gdn34
Bloodhound.RTF.12
Bloodhound.RTF.20
Exp.CVE-2017-11882!g5
Exp.CVE-2017-11882!g6
ISB.Downloader!gen60
ISB.Downloader!gen68
Web.Reputation.1
Trojan.Gen.MBT
Machine Learning-based
Heur.AdvML.B
Heur.AdvML.B!100
Heur.AdvML.B!200
Behavior-based
SONAR.TCP!gen1
SONAR.SuspBeh!gen667
Network-based
System Infected: Bad Reputation Process Request 4
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers.
Click here to learn more about Symantec's Endpoint Security Service.
Click here to learn how Symantec behavioral security technologies provide protection against zero-day attacks.
Click here to learn more about how Symantec Endpoint Protection uses advanced machine learning.
Click here to learn more about managing Symantec's Endpoint Security Intrusion Prevention System (IPS).
Click here to learn more about Symantec's Cloud Email Security Service. |
Protection Highlight: Lokibot unpacks into a brick wall
We've posted about the Lokibot infostealer multiple times over the years - as recently as Dec 14. Dating back to at least 2015, Lokibot spreads via a seemingly never-ending stream of spam campaigns. Definitely quantity over quality with these guys. Often leveraging malicious PDFs, RFTs and Office documents attached to the email as the infection vector and using common garden-variety quotes, shipping, banking, SWIFT, invoice and payment related social engineering themes, Lokibot attempts to steal credentials from hundreds of applications including browsers, FTP clients, Email clients, SSH clients, cryptocurrency wallets and password management software. Lokibot may use one of several different packing methods for obfuscation, but ultimately has to unpack itself into memory prior to execution of the main payload thereby exposing itself. At that point as the saying goes, The Emperor has no clothes.
Symantec continues to block Lokibot infection attempts with a wide variety of protection technologies. Defense in depth isn't just a concept, it really is the only way to protect your organization. The bad guys are out there all day every day looking for weak spots in your defense, and even relatively talentless attackers - through sheer persistence combined with a healthy dose of blind luck - will eventually find a hole in one or more of your security controls. But to find a way through multiple layers is a significantly more challenging task. And the best layer of all is of course, proactive defense where the protection is firmly in place well before the attack occurs. Looking at one of our telemetry monitoring systems over the past couple of weeks clearly demonstrates multiple Lokibot spam runs proactively blocked by one of many generic protections Symantec has in place against Lokibot and other threats.
Symantec has 0-day protection in place, identified by the following:
File-based
Packed.NSISPacker!g14
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers.
The file-based detection listed applies not only to our Email Security Service (ESS) but to all Symantec products that leverage our file-based protection technology, including Symantec Endpoint Protection (SEP), Data Center Security (DCS), Storage Protection, Server Protection, Web Security and Secure Web Gateway (SWG) among others.
To learn more about Symantec's Cloud Email Security Service, click here. |
Protection Highlight: Malvertising with JavaScript
JavaScript is a scripting language for creating dynamic webpage content. Wikipedia tells us that as of 2023, 98.7% of websites use JavaScript on the client side for webpage behavior. So if you're viewing a webpage, chances are it contains JavaScript within its code.
Malvertising is defined as the use of online advertising to spread malware, typically by injecting malicious or malware-laden advertisements into legitimate online advertising networks and webpages. These ads, containing the malicious code (which may or may not be written using JavaScript), may link directly to malware, or might end up redirecting users to another site (or even a chain of sites) which ultimately loads the final payload which may be an exploit kit, a phishing page, ransomware, etc.
Malicious JavaScript can be very difficult to spot, because it's essentially just text. When it's surreptitiously injected into websites, it can sometimes remain within the webpage for months or even years, including on very popular, well-known and entirely legitimate websites. And when people visit such a website and their security product suddenly pops open a scary looking window stating the website is infected, they could be forgiven for thinking the security product got it wrong, that it must be a "false positive" detection because the website is after all highly trusted by them.
About a year ago, Symantec received an inquiry along these lines - that browsing to a trusted website had produced a detection, a script file had been sent to quarantine, and was it justified? Analysis at the time showed the file was indeed bad, a malicious JavaScript redirector. Then just a few days ago we received another inquiry along the same lines, the same file but this time on a different website. Another review confirmed the original determination - malicious.
Although fortunately not a common occurrence, there's nothing overly unusual about this scenario. What was interesting, was looking up the JavaScript file on VirusTotal to see what sort of coverage it had. Apart from Symantec, only one other security vendor blocked it. Despite thousands of instances of the malicious script file in question floating around the Internet for years, clearly having infected a huge number of websites, and potentially a large number of people visiting those websites.
This is testament to how cunning threat actors are, and just how difficult it can be to detect some of the tricks they employ in their relentless efforts to compromise you and your valuable information.
Symantec has 0-day protection in place against this threat, identified by the following:
File-based
JS.Wonode
Click here to learn more about Symantec's Endpoint Security Service. |
Protection Highlight: Millions of Log4j attacks blocked by IPS
It's no secret that there are a multitude of vulnerabilities which have long been patched but are still widely-used by groups and individuals around the world. While the majority have been the most impactful and damaging the year they were disclosed, some continue to be highly effective years later - for example Log4j RCE CVE-2021-44228.
This vulnerability, disclosed in 2021 and widely known as "Log4Shell," is a critical security flaw that surfaced in the Apache Log4j 2 library, a popular Java-based logging utility. It enables remote attackers to execute arbitrary code with the privileges of the exploited application, posing a significant risk to systems using the affected Log4j version.
Symantec continues to observe threat actors around the world attempting to leverage this vulnerability on a daily basis, as you can see in the IPS telemetry chart below (note Symantec blocks these attacks as "Log4j2"). While we do have other protections in place for this vulnerability including anti-virus and DCS policies, our IPS technology is better posed to counter it due to the nature of the attack and this post only covers our network-based protections.
Plenty of information circulates regarding why threat actors persist in incorporating old and patched vulnerabilities into their arsenals. Some key reasons inclue the following:
Legacy Systems: Organizations grapple with outdated systems that are challenging to upgrade, leaving them exposed despite available patches.
Information Gap: Not everyone is promptly informed or takes action. The gap between vulnerability awareness and implementation allows attackers to exploit known weaknesses.
Resource Constraints: Limited resources, especially in smaller entities, hinder regular updates, making them susceptible to exploitation.
Adversarial Strategy: Attackers strategically target well-known vulnerabilities, capitalizing on the predictability that not all systems will be promptly patched.
Balancing Act: System administrators walk a tightrope between security and operational stability, sometimes delaying updates to avoid disruptions.
Human Inertia: Resistance to change, a common human trait, contributes to the persistence of exploiting familiar vulnerabilities despite known risks.
Symantec has 0-day protection in place against these Log4j attacks, identified by the following:
Network-based
Attack: Log4j2 RCE CVE-2021-44228 2
Attack: Log4j2 RCE CVE-2021-44228 4
Click here to learn more about managing Symantec's Endpoint Security Intrusion Prevention System (IPS). |
Protection Highlight: Mobile malware targets Japan
Over the past few months there have been regular campaigns targeting Japanese mobile users, including both consumers and enterprises, with malware for Android devices and phishing for iPhones.
The method of operation remains unchanged, with the attacker continuing to use malicious SMS as a means of infection along with sub-domains leveraging the free dynamic DNS hosted service, Duck DNS. Android users who are lured by these SMS messages are redirected to fake Softbank and KDDI websites and prompted to download fake security apps. iPhone users are redirected to a fake billing website. Here's a look at the attack pattern over the past month with each spike representing a new campaign.
In both cases (Android and iPhone) the exact domain redirected to depends on the particular mobile device in question, suggesting a higher degree of sophistication than is often the case in these campaigns. Below are some examples of the SMS messages:
* 【国税庁】重要なお知らせ、必ずお読みください
* 【国税庁 18:30】未払い税金お支払いのお願い。詳細はこちら【TT499】
* 【重要なお知らせ】未払い税金お支払いのお願い。ご確認ください【TQ177】
* 【重要なお知らせ HD474】SoftBank未払い料金お支払いのお願い
* 【12月6日利用停止予告】SoftBank未払い料金お支払いのお願い
* 【要確認】SoftBank重要なお知らせ、必ずお読みください
* 【ソフトバンク】お支払期限を過ぎた利用料金があります〔12月9日〕
* 【利用停止予告】KDDI未払い料金お支払いのお願い
A new wave that began earlier this month involved 757 Duck DNS sub-domains using the pattern of [10randomletters][.]duckdns[.]org. The Android malware, disguised as security software from Softbank and KDDI has the following capabilities:
Replacing the default SMS app
Collecting contacts, installed apps, phone numbers, and SMS messages and sending them to a remote server
Sending SMS messages without the user's consent
Modifying or deleting SMS messages if a certain pattern is detected
In essence, this malware turns the victim's phone into a bot, allowing it to further spread the infection and/or conduct other smishing campaigns involving the victim's contacts and other collected phone numbers. In addition, actors would have access to sensitive 2FA codes sent via SMS. This is harmful to both consumers and enterprises alike.
Japan is known for being a technologically advanced country with a large number of smartphone users. This makes it an attractive target for malicious actors looking to compromise mobile devices in order to steal sensitive information and further spread their malware. These type of attacks are only likely to increase.
Symantec has protection in place, identified by the following:
Mobile-based
AdLibrary:Generisk
AppRisk:Generisk
Web-based
Observed domains/IPs are covered under security categories.
To learn more about Symantec Mobile Endpoint Protection, click here. |
Protection Highlight: Mobile Phishing Attacks On The Rise
Mobile Phishing Attacks: Rising Trends and Urgent Solutions
Mobile protection remains a critical concern in today’s threat landscape. Despite industry efforts, SMS phishing attacks continue to proliferate. Recent actions by the FCC and carriers like T-Mobile aimed at combating scam texting in 2024 have not been sufficient. Our analysis reveals a troubling trend: SMS phishing attacks are not only prevalent but on the rise.
The Numbers
In February 2024, we meticulously analyzed millions of SMS messages, specifically focusing on potential malicious URLs. Shockingly, 12 out of every 1,000 messages (1.21%) were categorized as SMS phishing attacks. This represents a staggering 105% increase from the previous year and a 22% uptick from January 2024.
The Alarming Reality
The mobile landscape faces an evolving threat: attackers now target employees by creating realistic fake authentication portals. Their goal? To harvest valuable credentials and multi-factor authentication (MFA) codes. The attack chain often begins innocuously with an SMS phishing message sent to employee devices. This message contains a link to an Okta login page, enticing users to enter their account credentials and MFA codes.
Our observations indicate multiple attacks targeting users within organizations. We actively scan SMS messages for malicious URLs and promptly alert users about potentially harmful content. Additionally, we’ve witnessed credential phishing attacks via other communication channels, such as WhatsApp.
The Solution: Symantec Mobile Threat Defense (MTD)
When Enterprise Mobility Management (EMM) alone falls short, Symantec Mobile Threat Defense steps in. By rigorously checking URLs found in text messages against the threat intelligence in Symantec WebPulse, MTD provides timely and robust protection, effectively blocking users from accessing phishing sites on their mobile devices. MTD acts decisively, disrupting the attack chain right from the outset and shielding users from these malicious SMS phishing attempts.
Click here to learn more about Detection and Protection against malicious SMS messages.
Click here to learn more about Symantec Mobile Endpoint Protection.
Click here to learn more about Symantec's cloud-based Web Security Engine (WebPulse). |
Protection Highlight: New malware target old vulnerabilities
'You can't teach an old dog new tricks' is a proverb that means people are often reluctant or unwilling to change old habits. But you can teach a new dog old tricks and this is obviously not lost on at least some malware authors. CVE-2017-11882 is an aging Microsoft Office vulnerability but the 50,000 plus emails our security products blocked over the past month attempting to exploit it are new, and clearly demonstrates that attackers still believe it has value even years after a patch was released for it.
Not limited to email attacks as the initial vector of infection, CVE-2017-11882 is still being widely used by multiple groups and individuals to deploy various types of malware, with successful exploitation leading to the download and execution of a wide range of threats including Agent Tesla, WSH-RAT, Warzone RAT, Lokibot and Formbook, among others.
Symantec has 0-day protection in place for these new attacks, identified by the following:
File-based
AVE/Exp.CVE-2017-11882
AVE/Exp.CVE-2017-11882!g*
Exploit/CVE-2017-11882
AVE/Trojan.Mdropper
AVE/Trojan.Gen.NPE
W97M/Generic
* This indicates multiple similarly named detections exist such as AVE/Exp.CVE-2017-11882!g3, AVE/Exp.CVE-2017-11882!g7, etc.
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers.
To learn more about Symantec's Cloud Email Security Service, click here.
To learn more about Symantec's Endpoint Security Service, click here. |
Protection Highlight: Nokoyawa ransomware
Nokoyawa Ransomware has been actively operating since early 2022 and garnered attention for exploiting CVE-2023-28252, a vulnerability in the Windows Common Log File System Driver. Over time, the creators of Nokoyawa made significant changes to the ransomware, transitioning it to Rust.
Like other prominent ransomware actors, the authors employ a series of tactics, techniques, and procedures in their attack chain. In late 2022, they were observed utilizing IcedID's victims to gain initial access.
Upon successful compromise, encrypted files will be appended with a .NOKONOKO extension, and a ransom note (NOKONOKO-readme.txt or NOKONOKO-readme.html in latest iterations) dropped on the machine. Below is an example:
The primary targets of Nokoyawa ransomware are businesses and organizations, with a specific focus on industries such as healthcare, financial services, government, and manufacturing. These sectors are often chosen due to their possession of sensitive data, critical operations, and willingness to potentially pay significant ransoms. Reports suggest that ransoms typically average around 200,000 U.S. dollars worth of Bitcoin.
These are some of the MITRE TTPs used by Nokoyawa's actors:
Command and Scripting Interpreter: Windows Command Shell [T1059.003]
Windows Management Instrumentation [T1047]
Data Encrypted for Impact [T1486]
Impair Defenses: Disable or Modify Tools Defacement [T1491]
Defacement [T1491]
Symantec has 0-day protection in place against this threat, identified by the following:
Behavior-based
SONAR.WMIC!gen12
SONAR.RansomPlay!gen1
SONAR.Ransom!gen35
SONAR.Ransomware!g13
MEMSCAN.Ransom!gen1
MEMSCAN.Ransom!gen8
SONAR.SuspLaunch!g258
ACM.Wmip-Ps!g1
MEMSCAN.Ransom!gen2
SONAR.Ransom!gen98
SONAR.RansomGen!gen3
SONAR.RansomNoko!g3
EDR-based
Symantec EDR is capable of monitoring and will flag this threat actor's tactics, techniques and procedures
Symantec added malware specific threat hunting queries which can be fired on customer iCDM console. For more information, you can refer to this link: https://github.com/Symantec/threathunters/tree/main/Ransomwares/Nokoyawa.
File-based
Ransom.Nokoyawa
Machine Learning-based
Heur.AdvML.B
Policy-based
Symantec Data Center Security out-of-the-box default hardening policies for Windows provide 0-day protection against unknown threats, including previously unseen ransomware variants and associated behaviors.
To learn how Symantec behavioral security technologies provide protection against zero-day attacks, click here.
To learn more about Symantec's Endpoint Security Service, click here.
To learn more about how Symantec Endpoint Protection uses advanced machine learning, click here.
To learn more about Symantec's Data Center Security, click here. |
Protection Highlight: Obfuscated phishing attack no match for Stargate
Phishing - a type of social engineering attack used to trick a person into revealing sensitive information to the attacker for the purpose of stealing confidential data including login credentials and credit card numbers - can be very difficult for people to spot, and potentially quite damaging to the victim. One of the most common forms of cybercriminal activity, particularly against business enterprises, phishing attacks significantly increased in 2022. There is a huge variety of phishing kits available on the underground market and Symantec blocks new phishing attacks on a daily basis. Last week was no exception.
On December 5 one of our monitoring systems alerted to a spike in malicious email traffic.
A quick investigation showed our Stargate security engine (leveraged by our Email Security Service or ESS) with it's advanced heuristic threat detection capabilities, had proactively blocked a sneaky email attack where obfuscated code was hiding (or in this case, attempting to hide) waiting to spring into action. Once de-obfuscated and executed, the code targets HTML in a phishing attempt to capture and steal Microsoft Office 365 login credentials and passwords. If the victim is successfully lured the stolen information will be uploaded to a CnC server controlled by the attackers. Game over for the victim.
Given the great lengths the authors go to in their attempts to hide the infection vector in order to avoid detection, coupled with the often entirely legitimate looking "phish" page or pop-up, it can be next to impossible for the unsuspecting target of the attack to detect. Fortunately, Symantec ESS customers didn't need to, as the advanced heuristics engine within the Stargate security stack did it for them. Game over for the attacker.
Symantec has 0-day protection in place, identified by the following:
File-based
Scr.Malscript!gen2
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers.
To learn more about Symantec's Cloud Email Security Service, click here.
To learn more about Stargate - a threat detection platform based on machine learning, cloud knowledge, and deep content inspection - contact Symantec here. |
Protection Highlight: PDF attacks on the rise
Over the past few months we've observed a surge in PDF attacks. In order to combat these attacks, Symantec's innovative PDF heuristic solution leverages advanced heuristic and machine learning techniques. This proactive approach has proven highly effective, successfully blocking numerous attacks including those orchestrated by notorious threat actors such as TA544, TA577 and RogueRadicate.
Below are some of the spam campaigns blocked by our heuristic solution during January and February where threat actors were attempting to leverage PDFs in their attack chains:
Jan 9 ~ spam run with PDF attachment delivers Netsupport RAT as final payload
Jan 16 ~ spam run with PDF attachment delivers Lokibot as final payload
Jan 16, 24 & Feb 1, 2 ~ spam run with PDF attachment delivers DBatLoader and Remcos as final payload
Jan 17, 31 ~ spam run with PDF attachment delivers Wikiloader as final payload
Jan 24, 25, 26, 29 and Feb 12, 13 ~ spam run with PDF attachment delivers Darkgate as final payload
A snapshot of malicious PDF blocks show sizeable spikes in late January and again mid February
NetSupport RAT
It was observed that the recent campaign delivering NetSupport RAT used similar PDFs to the Darkgate campaigns in November last year. Therefore, it is believed that they are sent by the same threat actor, RogueRadicate. It also drops the malicious payload when the user clicks on the Download link in the PDF.
Lokibot
Worthy of attention is a trend where an increasing number of Infostealers are opting to distribute via PDF spam campaigns. We captured a wave of Lokibot PDF campaigns in January this year where an office file is dropped when the user opens the PDF attachment. The final payload is usually downloaded and executed by the dropped office file via well-known vulnerabilities like CVE-2017-11882 or CVE-2017-0199.
DBatLoader & Remcos
We also recenty observed some PDF spam campaigns that intermittently deliver DBatLoader & Remcos as their final payload. The email pretends to be an invoice but requires the user to update the Adobe Acrobat software in order to view the full document. If the user clicks the update link, the threat will download the malicious payload instead of the software update.
Wikiloader
We covered the return of the Wikiloader spam campaign this year in a previous Protection Highlight. The PDFs used in the spam run pretends to be logistic company invoices for download. Upon clicking the link, the user will be redirected to a website where a malicious payload is dropped.
Darkgate
Darkgate exhibited significant activity from October to November of last year. Following a period of relative silence in December, they reappeared in January. The PDFs employed in their spam tactics display alerts such as 'Access Denied' or 'Document Not Displayed Correctly,' aiming to persuade users to click a link for either a purported Adobe Reader update or to download the document for offline viewing. As with the others, the link redirects users to a website where a malicious payload is dropped. For additional information about the Darkgate activity in January 2024, see the following Protection Bulletin.
One more thing...
Besides mitigating the above spam campaign attack waves, the PDF heuristic solution effectively blocks phishing PDFs distributed via email attachments or links, which aim to extract users' confidential information such as credit card details, banking account credentials, or email account login information.
Symantec protects you from these PDF threats, identified by the following:
File-based
Scr.DLHeur!gen1
Scr.DLHeur!gen2
Scr.DLHeur!gen3
Scr.DLHeur!gen5
Scr.DLHeur!gen6
Scr.DLHeur!gen7
Scr.DLHeur!gen8
Scr.DLHeur!gen9
Scr.DLHeur!gen10
Scr.DLHeur!gen13
Scr.DLHeur!gen14
Scr.DLHeur!gen15
Trojan.DLHeur!gen2
Trojan.DLHeur!gen3
Trojan.DLHeur!gen4
Trojan.DLHeur!gen5
Phish.Pdf!gen2
Phish.Pdf!gen3
Phish.Pdf!gen4
Phish.Pdf!gen5
Phish.Pdf!gen6
Phish.Pdf!gen7
Phish.Pdf!gen8
Scr.Qbot!gen12
Scr.Qbot!gen14
Scr.Qbot!gen19
Email-based
Coverage is in place for Symantec's email security products
Click here to learn more about Symantec's Endpoint Security Service.
Click here to learn more about Symantec's Cloud Email Security Service. |
Protection Highlight: Phishing on the rise
Phishing is a type of social engineering attack where the attacker sends spam messages - usually email or text but increasingly also including social media and even phone calls - designed to fool the target into revealing sensitive information such as credit card numbers or login credentials, or to install malware on the victim's device. Phishing attacks are increasing, and anyone who uses email, text messaging and other forms of communication (i.e. most of us) is a potential victim.
Statistics show billions upon billions of spam emails & texts are sent on a daily basis, and while some surveys have the percentage of spam at around 50% of all email, others report it being well over 80%. Suffice to say, it's a significant percentage resulting in a sizeable number of unsolicited and unwanted messages. Other spam-related statistics include reports that over 80% of all companies experience at least one phishing attack every year, with a similar percentage seeing a year-on-year increase in attacks. The FBI reported over 300,000 phishing attacks in 2022 alone. It's been reported that a typical phishing attack on a company costs close to 5 million dollars to resolve. Sobering stuff. Yet another survey reported around 90% of corporate breaches were the result of phishing attacks.
Some of the most common phishing attacks an organization may face include:
Email Phishing - the most common form of phishing where the attacker sends a deceptive email that appears to be from a legitimate source.
Spear Phishing - a more targeted form of attack where the attacker has researched their target in order to tailor their message.
Whaling - these attacks target high-profile individuals, such as company executives.
Pharming - this involves redirecting victims to a fraudulent website that mirrors a legitimate website.
Common senders include:
Delivery companies
Banking & Financial institutions
Ecommerce companies
Govt services
Recruitment services
In December of last year we posted about Stargate being equal to the task of stopping obfuscated phishing attacks in their tracks. At the time our telemetry systems were recording an average 2,500 blocks per day with a sizeable spike observed on December 5th. In contrast, we now see an average of around 17,000 blocks per day with spikes exceeding 30,000. It doesn't appear on the chart below but early August saw a spike of over 60,000 blocks. It doesn't look like phishing is going away anytime soon.
Symantec has 0-day protection in place against this threat, identified by the following:
File-based
Scr.Malscript!gen2
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers.
Web-based
Observed domains/IPs are covered under security categories in all WebPulse enabled products.
Click here to learn more about Symantec's Endpoint Security Service.
Click here to learn more about Symantec's Cloud Email Security Service.
Click here to learn more about Symantec's cloud-based Web Security Service (WebPulse).
To learn more about the Stargate Security Service (a threat detection platform based on machine learning, cloud knowledge, and deep content inspection) contact Symantec here. |
Protection Highlight: Play Ransomware
First seen around mid 2022, Play ransomware (also known as PlayCrypt) is among the most active ransomware in today's threat landscape, going toe to toe with other notorious variants such as LockBit, Mallox, Clop, etc. Over the past few weeks, the gang has claimed more than 25 victims, targeting both private and public organizations of all sizes and industry types. Like its notorious peers, this actor is known for its use of the double extortion tactic where victims are threatened with having their data sold (on top of already having it encrypted) if they do not pay the ransom.
When it comes to the infection vector, the threat actors behind Play ransomware have been using known vulnerabilities (ProxyNotShell being one example) and have bought infrastructure access in the form of stolen credentials from other threat actors who have already managed to compromise their targets. The actors also use various tools such as Cobalt Strike, MimiKatz, Empire and remote access trojans (RATs) for lateral movement and persistence. Once installed on a system, Play ransomware will encrypt all files with the .PLAY extension and create a ransom note named "ReadMe.txt" which contains instructions on how to pay the ransom. The ransom note is known to be short, often containing only the word "Play" along with a link to the attacker's onion website, or an email address with which to contact them.
Symantec has 0-day protection in place against this threat, identified by the following:
Behavior-based
SONAR.RansomPlay!gen1
SONAR.RansomPlay!gen2
SONAR.RansomPlay!gen3
File-based
Ransom.PlayCrypt
Ransom.PlayCrypt!g1
Ransom.PlayCrypt!g2
Machine Learning-based
Heur.AdvML.B
Heur.AdvML.B!200
Policy-based
Symantec Data Center Security out-of-the-box default hardening policies for Windows provide 0-day protection against unknown threats, including previously unseen ransomware variants and associated behaviors.
To learn how Symantec behavioral security technologies provide protection against zero-day attacks, click here.
To learn more about Symantec's Endpoint Security Service, click here.
To learn more about how Symantec Endpoint Protection uses advanced machine learning, click here.
To learn more about Symantec's Data Center Security, click here. |
Protection Highlight: Replit abused in South Korean phishing campaigns
Replit is an online platform that provides an integrated development environment (IDE) for various programming languages. It also offers a feature called "Repl.it Sites" that allows users to host static websites directly from their Replit accounts. Users can create HTML, CSS, and JavaScript files to build their websites and then deploy them with just a few clicks.
Symantec continues to observe actors abusing Replit and hosting their phishing webpage on the platform. It bears similarities to the way people abused CloudFlare R2 and IPFS, albeit less widespread but nonetheless still dangerous.
Over the past few weeks, we have blocked an actor who has been running multiple Replit phishing campaigns in South Korea, targeting both local and international companies in the country and starting out with what appears to be small scale testing before ramping up on December 6.
If users are successfully tricked by the malicious emails and happen to click on the URL provided (example: hxxps[:]//view-file--kisiy37297[.]repl[.]co/#[recipient email address]), they'll end up on a phishing page hosted on Replit, which is actually a fake MailPlug (a popular webmail service in South Korea) login page.
Observed email subjects:
[KOTRA인천지원단] 사업제안
[주식회사 켄드리드마트]로부터 [주문서]가 도착했습니다.
(주)썬패치테크노에서 주문서를 보냈습니다.
Symantec has 0-day protection in place against IPFS attacks, identified by the following:
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers.
Web-based
Observed domains/IPs are covered under security categories. Newly observed URLs are detected and categorized using known Replit URL patterns.
Click here to learn more about Symantec's Cloud Email Security Service.
Click here to learn more about Symantec's cloud-based Web Security Engine (WebPulse). |
Protection Highlight: SEP IPS vs RDP Attacks
Remote Desktop Protocol (RDP) is a feature of Microsoft Windows that provides remote access. Remote workers use it to connect to computers physically located at their office and IT experts can use it to fix someone's computer from across the world. It was already popular prior to COVID-19, but since the Pandemic began and huge numbers of company employees suddenly found themselves working from home, RDP connections grew exponentially. Unfortunately it's just as convenient for those with ill intent, and hackers continuously conduct RDP attacks in order to access and compromise enterprise networks.
An RDP attack is a type of cyberattack that attempts to gain access to a remote computer using the RDP protocol. These attacks are a very popular method for hackers to take advantage of insecure systems, exposed services, and vulnerable network endpoints. A successful RDP attack may allow an attacker to harvest credentials, execute malicious code, or even give them full control over the target system.
Symantec blocks hundreds of thousands of RDP attacks every week.
Alarmingly, RDP attacks are increasingly being used by cybercriminals and nation-state actors to carry out ransomware attacks. By gaining access to a victim's computer through RDP, attackers can install ransomware that encrypts the victim's files and demands payment for the decryption key.
There are a variety of tools and techniques that actors use to perform RDP attacks.
* Scanning tools: Attackers may use scanning tools to search for RDP servers that are connected to the internet and are vulnerable to exploitation. These tools can help attackers identify potential targets for RDP attacks.
* Brute-force tools: Brute-force tools are used to attempt to crack passwords by trying many different combinations until the correct one is found. Attackers can use these tools to try to gain access to RDP connections that are protected by weak or easily guessable passwords.
* Exploit kits: These are collections of tools and exploits that are used to identify and exploit vulnerabilities in software. Some exploit kits are specifically designed to target RDP vulnerabilities.
* Credential-stealing malware: Attackers may use malware that is designed to steal login credentials from victims' computers. This type of malware can be used to steal RDP login credentials, as well as other login credentials that are stored on the victim's computer.
* Social engineering: Attackers may use social engineering techniques, such as phishing emails, to trick victims into revealing their login credentials or downloading malware that can be used to perform RDP attacks.
RDP attacks pose a significant threat and should be taken seriously by individuals and organizations alike. It's important to ensure that RDP connections are properly secured with strong passwords, two-factor authentication, and other security measures.
Symantec has 0-day protection in place against RDP attacks, identified by the following:
Network-based blocking signatures
Attack: NCrack Tool RDP BruteForce Activity
OS Attack: Microsoft Windows Remote Desktop Services RCE CVE-2019-0708*
OS Attack: RDP Scan Attempt 2
Web Attack: Microsoft RDP Exploit Attempt
Attack: Microsoft RDP CVE-2012-0002 4
System Infected: GoldBrute RDP BruteForce Attempt
* This indicates multiple similarly named detections exist.
To learn more about managing Symantec's Endpoint Security Intrusion Prevention System (IPS), click here.
To learn more about Symantec's Endpoint Security Service, click here. |
Protection Highlight: SEP Mobile leverages WebPulse to block Smishing attacks
Smishing, also known as SMS phishing, involves cybercriminals using SMS (Short Message Service) text messages to deceive mobile device users and trick them into divulging sensitive information, clicking on malicious links, or downloading harmful attachments.
Smishing activities have been steadily increasing over the past few years primarily due to the widespread adoption of smartphones and other mobile devices among both consumer and enterprise users. Other factors have also played a crucial roles in its growth:
Interconnectivity of web services with mobile phones: Many apps and services require users to provide their mobile phone numbers for verification or authentication purposes. This interconnectivity between web services and mobile phones provides attackers with an additional means of obtaining users' phone numbers and crafting personalized smishing messages.
Lowering effectiveness of email phishing: Email providers and security solutions have become better at detecting and blocking email phishing attempts. As a result, some attackers have shifted their focus to SMS as a relatively untapped avenue for phishing attacks.
Lack of user awareness: Many people are not as cautious with text messages as they might be with emails, making them more susceptible to falling for smishing attempts.
URL shorteners: Attackers often use URL shorteners to disguise malicious links in SMS messages, making it harder for recipients to discern their true destination via mobile devices.
Difficulties in tracking and attribution: Unlike email phishing, where headers and sender information can sometimes be traced, SMS-based attacks can be more challenging to trace back to their source, providing some level of anonymity to attackers.
While SMS phishing initially targeted consumers, enterprise users have increasingly become prime targets for threat actors due to the potential rewards from obtaining corporate information. Enterprise users often have access to sensitive and valuable corporate data, including intellectual property, financial data, customer databases, and proprietary technologies.
Over the past few months, SMS attacks have been on the rise. Symantec Endpoint Protection Mobile has blocked more than 50,000 malicious SMS messages sent to corporate mobile users (including both Android and iOS) around the world.
Symantec Endpoint Protection Mobile is also able to analyze links contained within SMS messages. It shields users from SMS phishing attempts by checking URLs found in text messages against the threat intelligence in Symantec WebPulse, part of the Symantec Global Intelligence Network (GIN), and alerting users when the links are suspect.
To learn more about Symantec Mobile Endpoint Protection, click here.
To learn more about Symantec's cloud-based Web Security Service (WebPulse), click here.
To learn more about Symantec's Global Intelligence Network (GIN), click here. |
Protection Highlight: SEP Mobile MITM Protection
A modern-day man-in-the-middle (MITM) attack is a type of cyberattack where communications between two parties are intercepted and may be altered, while making it appear as if a normal exchange of information is underway. MITM attacks usually aim to steal information, but may also have other intent such as spying on victims, sabotaging communications, or corrupting data.
Mobile SSL MITM attacks are becoming increasingly prevalent as more and more people use mobile devices for online activities such as banking, shopping, and accessing sensitive data. These attacks are facilitated by the widespread use of public Wi-Fi networks which are often unsecured and vulnerable to interception, exploiting the SSL/TLS encryption used to secure communication between a mobile device and a server.
In a nutshell, a Mobile SSL MITM attack works like this:
* The attacker sets up a fake Wi-Fi hotspot or a rogue access point that mimics a legitimate hotspot
* The victim connects to the fake hotspot and begins browsing the internet or accessing online services
* When the victim accesses a secure website, the attacker intercepts the SSL/TLS traffic and establishes a fake SSL/TLS connection with the victim's mobile device
* The attacker acts as a proxy between the victim's device and the server, intercepting and decrypting the encrypted traffic
* The attacker can now read and modify the traffic passing between the victim's device and the server
* The attacker can steal sensitive information such as login credentials, credit card numbers, or personal data
* The attacker can then modify the traffic and inject malicious code or malware into the communication stream, further compromising the victim's device
The chart below shows the trend of SEP Mobile SSL MITM blocks for the year to date.
SEP Mobile works to thwart MITM attacks as follows:
Active Honeypot technology thoroughly analyzes network connections end-to-end to keep communications safe
When SEP Mobile detects a MITM network attack, two things happen simultaneously:
Our automated network protection response kicks in, including actions such as tunneling traffic from the mobile device through a secure VPN connection or blocking the mobile device from accessing sensitive corporate resources (email, work applications, etc.)
Both enterprise security administrators and end-users receive alerts about the threat, with the alert including details about certificates and network events involved in the attack
To learn more about Symantec Mobile Endpoint Protection, click here. |
Protection Highlight: Simulated Android DOS Attack
In our daily work we constantly monitor for anomalies and signs that may indicate a new or modified threat attempting to avoid detection. Our automation systems recently alerted us to an issue where our cloud scanner noticeably slowed when processing a select set of APK samples. A quick look showed several of the samples in question had also had a slight impact on the performance of the broader scanning automation system.
Other vendors on VirusTotal appear to have been even more adversely affected showing timeouts during analysis, effectively making this appear to be a mobile DOS attack.
A deeper look into the samples revealed the following attributes:
Each of the container samples contained a large number (up to 32K) of small files
These smaller files contained what look like random bytes, and didn’t appear to be used by the APK
The APKs themselves seem to be generated by automation
The package names also appeared to be random and automatically generated
* hbjvuozxxuhpyfbnvxtd.dxgbesgyxlsvkccgozbi.deomdyhhwgeraghfzrds
* dknmmohvktzcmopltehs.yrrnbpuvyctwiefeeycz.gtwxbuikrwakaehwpazg
Based on behavioral analysis, we determined that these APK samples are mostly Adware/Grayware, the general term for these and other potentially undesirable but not outright malicious programs being "Potentially Unwanted Applications" or PUA for short. Some of them connect to a remote server to get ad-related configurations or integrate with some of the many SDKs which manage notification push or location service.
We have seen over 11,000 of these samples (each containing over 30,000 files) to date. We believe they may be some sort of test - hence the word "simulated" in the title - but it's difficult to determine whether such a test was conducted by a security researcher or someone with malicious intent.
The ability to learn and adapt is paramount for a security company to be successful and we immediately modified our automation to allow a more efficient analysis of these samples, effectively overcoming the processing performance hit observed earlier.
To learn more about Symantec Mobile Endpoint Protection, click here. |
Protection Highlight: Snake Keylogger
Snake Keylogger, malware developed using .NET, has been around for several years now and is still highly prevalent. Designed to record and transmit keystrokes made on a computer to a remote server along with screenshots and clipboard data, it is used by multiple groups and individuals around the world in both targeted and non-targeted email spam campaigns of varying sizes.
The main method it uses to spread is identical to that used by two other highly prevalent infostealers, Formbook and Agent Tesla - emails containing malicious attachments such as Microsoft office documents or PDF files. The emails arrive with common social engineering subjects which include keywords such as Quote, Purchase Order, Invoice, Payment, SWIFT, Shipping, etc. The attachment is usually an archive which when unpacked and executed, may employ several evasion tactics in an attempt to avoid detection including embedding files, loading remotely-hosted exploits and encrypting shellcode before ultimately deploying its information stealing payload.
As the chart below shows, Snake Keylogger remains active in the threat landscape and as the protection list below the chart shows, Symantec has a multitude of different coverage in place. Our priority is protecting our customers by stopping threats and the best type of protection is one that catches a wide variety of different malware, as opposed to coverage that is tailored to a specific threat and only stops that one particular attack (although we have one of those too --> Trojan.SnakeKeylogger).
Actually only a subset of the total blocks, but the spikes clearly demonstrate new campaigns.
Symantec has multiple 0-day protections in place against this threat:
File-based
Scr.Malcode!gdn*
Scr.Malcode!gen*
Packed.NsisPacker!g*
Msil.Packed.*
Bloodhound.RTF.*
WS.Malware.*
WS.SecurityRisk.*
Trojan.Gen.*
Trojan Horse
Trojan.SnakeKeylogger
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers.
Machine Learning-based
Heur.AdvML.*
* This indicates multiple similarly named detections exist such as Scr.Malcode!gen25, Scr.Malcode!gen34, etc.
To learn more about Symantec's Endpoint Security Service, click here.
To learn more about Symantec's Cloud Email Security Service, click here.
To learn more about how Symantec Endpoint Protection uses advanced machine learning, click here. |
Protection Highlight: South Korea targeted by Reg Stealer
A registry file (.reg) is a text-based configuration file used in Microsoft Windows operating systems to modify the Windows Registry, a hierarchical database that stores configuration settings and options for the operating system, hardware, software, and user preferences.
While .reg files are typically used for legitimate purposes, cybercriminals can exploit them to facilitate the download and execution of malware on a victim's system. In today's threat landscape, while not exactly prevalent, they are still being actively used by certain groups and individuals around the world in their attack chains.
Recently, Symantec observed a malspam campaign targeting organizations in South Korea (both local and international). Attached to the malicious email (subject: "확인 필요: 송장 처리를 위한 지불 세부 정보") are the same two malicious .reg files "PO_Inv.XLS.reg" and "Beneficiary Details.reg". If executed successfully, these files will download malicious PowerShell and bat files.
The bat file (ld.bat) will execute the PowerShell file (ld.ps1), and subsequently, the PowerShell script will attempt to modify the machine's registry and the %APPDATA% folder, exempting it from being scanned by Windows Defender. It will also download an infostealer posing as PuTTY, an open-source terminal emulator. All downloaded files will be stored in the %APPDATA% folder.
This Python-based infostealer, named 'REG STEALER 2023,' is highly obfuscated and encrypted. It shares numerous similarities with various Discord and Telegram stealers. Upon analysis, we believe that this stealer is very likely based on Blank Grabber's source code which had been posted on a popular web-based platform for version control and collaborative software development. Below are some of the data it can capture from compromised machines and send through Discord Webhooks and Telegram Bots.
Browsers passwords, cookies, autofills and history
Discord tokens
Saved WIFI passwords
System info
Screenshot
Telegram session
Common files
Crypto Wallets
Webcam
Various games and gaming platform's cookies and sessions
Symantec protects you from this threat, identified by the following:
Email-based
Coverage is in place for Symantec’s email security products
File-based
Downloader
Trojan Horse
Trojan.Regsteal
Web.Reputation.1
Machine Learning-based
Heur.AdvML.B
Web-based
Observed domains/IPs are covered under security categories in all WebPulse enabled products
Click here to learn more about Symantec's Endpoint Security Service.
Click here to learn more about how Symantec Endpoint Protection uses advanced machine learning.
Click here to learn more about Symantec's Cloud Email Security Service.
Click here to learn more about Symantec's cloud-based Web Security Service (WebPulse). |
Protection Highlight: Symantec Advanced Machine Learning Technology
How Symantec Uses Advanced Machine Learning Technology To Protect Against Zero-Day Threats
Machine Learning, often referred to as just ML, is a signatureless technology that can block new malware variants at the pre-execution stage. At Broadcom, ML is used at various layers to protect our customers from cyberthreats. These layers have been designed to both proactively and reactively "gate" every location a suspicious file, OS event, registry entry, URL or network activity is seen by our products - including Endpoints, Gateways and in our backend analytics platform. Symantec has the ability to dynamically analyze new content as soon as it's available with a comprehensive set of threat scanning engines, feeding the data into the Symantec Global Intelligence Network (GIN). Symantec uses security telemetry from millions of endpoints, threat-related data feeds from third-party security vendors, along with a rich set of clean files to train and evaluate various ML models. These models are deployed at numerous products to detect threats, both on customer endpoints as part of our agent and in our backend analysis systems.
Zero-day Protection is Critical
In addition to the analytics platform mentioned above, we also leverage a Cloud Sandbox Analysis Engine (rather aptly named "Cynic") running multiple ML models and Clustering algorithms to classify and cluster files according to their threat type, risk potential, dynamic and static metadata, and behavior. Symantec analyzes customer submissions using both automated systems and human malware analysts as quickly as possible, and the intelligence is feed into ML training models for improving classification efficacy. Our multi-model Advance Machine Learning technology runs on various file types, in both 32-bit and 64-bit avatars to provide actionable analysis. Where gaps in protection are identified, they are analyzed by backend ML models and blocked through Reputation lookups without delay. The primary objective of Symantec Advanced Machine Learning is to protect against new and unknown malware, commonly known throughout the industry as Zero-day attacks. This is where ML excels.
In the last quarter alone, Symantec's Advanced Machine Learning blocked almost 23 Million threats on Symantec Endpoints and Gateway products. Around 3.9 Million of these blocks were against Zero-day attacks - i.e. never before seen by any of our security products or protection technologies. This is what is meant by "proactive" protection, as opposed to "reactive" which means adding new or updating existing protections in response to the attack. Proactive protection is the panacea against cyberthreats and the bane of would-be cybercriminals everywhere.
During the past quarter Symantec Advanced Machine Learning provided the following protection:
13.5M threats blocked by Symantec Advanced ML on our Gateway products
9.3M threats blocked on Endpoints
3.9M Zero-day threats were blocked by ML, including:
9K Ransomware (Cerber, Cryptodefence, Gandcrab, Ryuk, Wannacry, Zombie, etc.)
512K Trojan (Emotet, Cridex, Whispergate, etc.)
160K Win32 (Qakbot, Fujacks, Expiro, etc.)
230K Backdoors (Cobalt, Limitail, Berbew, etc.)
1.1M browser-based threats were blocked on Endpoints - 32% from Chromium, 24% from MSEdge and 15% from Firefox
731K threats launched through command line to download and execute malicious files were blocked on Endpoint products
585K threats were blocked attempting to enter the system from external sources such as USB drives
200K attacks blocked using SMB for network file sharing
105K threats blocked which were downloaded using peer to peer (P2P) networking programs such as Anydesk (RDP), Utorrent and Bittorrent
5.9K threats downloaded using Scripting host (Powershell/csript/wcript) blocked
Charting select Zero-day protections on Endpoints and Gateways over the quarter
Click here to learn more about how Symantec Endpoint Protection uses Advanced Machine Learning.
Click here to learn more about Symantec's Endpoint Security Service.
Click here to learn about Symantec's Cloud Sandbox Analysis Engine (Cynic). |
Protection Highlight: Targeted attacks on Middle East Governments
Amidst the ongoing conflicts in Eastern Europe and the Middle East, a malicious actor has been orchestrating cyber espionage campaigns against Middle Eastern governments. These campaigns are exploiting the heightened fears and tensions prevailing in the region due to the protracted conflicts. Such malicious activities only exacerbate disruptions for affected governments and their citizens, posing an even greater threat to the stability and security of the region.
The attack is initiated through sophisticated spear phishing emails, with the attackers impersonating a reputable US-based aerospace and defense company. These meticulously crafted emails leverage military and law enforcement products as a primary lure. In actuality, the attacker entices users to access an FAQ document they’ve fabricated. Within this document, detailed descriptions of each product are provided, in addition to their unique attributes, technical specifications, and recommended applications.
Should a user be successfully tricked into opening the attached ZIP file (FAQ.zip), they will encounter another ZIP file of the same name (FAQ.zip) containing an IMG archive (FAQ.img). This second archive hosts a PDF decoy (FAQ.pdf) alongside malicious LNK (FAQ.pdf.lnk), EXE (Graphic.exe), and DLL (d3d8.dll) files.
Upon execution, the malicious LNK file will activate the PDF decoy whilst executing the EXE file in the background. Additionally, the DLL (d3d8.dll) is loaded using a search order highjacking technique. This DLL then loads the decrypted backdoor code which initiates system fingerprinting and starts C2 registering of the host, assigning random digit IDs which wait on standby to receive and execute further commands.
Symantec protects you from this threat, identified by the following:
File-based
Backdoor.Trojan
Scr.Malarchive!gen1
Trojan Horse
Machine Learning-based
Heur.AdvML.C
Adaptive-based
ACM.Ps-Rd32!g1
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers.
Policy-based
Symantec Data Center Security out-of-the-box default hardening policies for Windows protect MS Office applications.
DCS prevents MS Office applications from launching command interpreters including cmd.exe, powershell.exe or winrar.exe / winzip.exe as child processes.
DCS will stop execution of any executable file further along the attack chain for this threat.
For additional protection, customers can configure DCS network rules to set up a network perimeter for privileged services and applications.
Click here to learn more about Symantec's Endpoint Security Service.
Click here to learn more about how Symantec Endpoint Protection uses Advanced Machine Learning.
Click here to learn more about Symantec's Endpoint Security's Adaptive Protection.
Click here to learn more about Symantec's Cloud Email Security Service.
Click here to learn more about Symantec's Data Center Security. |
Protection Highlight: Thanks a (VipersoftX) million WebPulse. Literally. Every day.
VipersoftX is a particularly sneaky piece of information stealing malware that's been around since at least late 2019. A heavily obfuscated and regularly updated RAT which spreads through cracked software via torrents and software-sharing sites, it has recently been observed deploying as small PowerShell scripts concealed inside large system log files in an attempt to avoid detection. It's main goal is to steal cryptocurrencies, employing techniques such as clipboard swapping, host machine fingerprinting, and the downloading of additional malicious payloads onto infected machines.
Newer versions of ViperSoftX contain a payload used to install a browser extension which effectively provides it with access to every page the victim visits, allowing it to carry out man-in-the-browser attacks to perform cryptocurrency address swapping and steal credentials along with other clipboard content. It steals cryptocurrencies by checking the content of the clipboard to see if it holds a pattern that matches a cryptocurrency wallet address. If it finds a match, it overwrites the clipboard data with its own wallet addresses. Simple but effective.
Symantec's WebPulse ecosystem has long been detecting and blocking high volumes of traffic from the malicious browser extension deployed by VipersoftX. On average, over a million URL requests per day.
The malicious browser plugin generates a large number of DGA-style fake domains that show up in our customers' Web traffic. Among its extensive arsenal, WebPulse has a number of generic "botnet traffic detectors" and some of these are automatically triggering on this traffic. Symantec customers can use these WebPulse detections (the category being returned is "Malicious Outbound Data / Botnets") in their logs to identify the machines in their network that are generating this traffic.
Symantec has 0-day protection in place against this malicious browser extension, identified by the following:
Web-based
Category: Malicious Outbound Data / Botnets
To learn more about Symantec's cloud-based Web Security Service (WebPulse), click here. |
Protection Highlight: UPS Phishing Attack Blocked
Phishing attacks, one of the most common forms of cybercriminal activity, significantly increased in 2022 and the pace doesn't appear to be slowing any in 2023. Symantec security products detect and block phishing campaigns all day every day, but occasionally the spike in malicious email traffic is significant enough to raise eyebrows, as it tends to indicate that someone is trying harder than usual to steal logins to some system or other. This past week our monitoring systems alerted us to a sizeable phishing campaign targeting UPS credentials with over 50,000 emails at it's peak, all proactively blocked by our Stargate security engine (which underlies our Email Security Service) with it's advanced heuristic threat detection capabilities.
The February 20-22 peak represents the UPS phishing campaign
The emails contain a malicious attachment which unpacks to an HTML file displaying a fake UPS login page entitled "UPS Worldwide Saver", and asking the victim to verify their identity by providing their login username and password in order to "prove You are Human". Entering the requested information will result in a POST action which uploads the stolen credentials to an attacker-controlled server. A successful compromise may also result in the login details being used for further attacks as many people unwisely use the same credentials for different online services. To help protect yourself against phishing attacks, never click on links that appear in unsolicited emails and instant messages, and never provide your personal information in response to an unsolicited request, whether over the phone or the Internet.
Symantec has 0-day protection in place, identified by the following:
File-based
Scr.Heuristic!gen2
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers.
To learn more about Symantec's Cloud Email Security Service, click here.
To learn more about Stargate - a threat detection platform based on machine learning, cloud knowledge, and deep content inspection - contact Symantec here. |
Protection Highlight: Ursnif Impersonates Italian IRS
Ursnif (aka Gozi, Snifula) is a well-known banking Trojan family that has been around for well over 15 years. It's source code has been leaked to the public several times over this period resulting in many different strains and variants. Delivered primarily through malicious spam campaigns, Ursnif is designed to steal login credentials, two-factor authentication codes, banking details and other confidential data from infected systems.
Recently observed campaigns distributing Ursnif have been targeting financial institutions based in Italy, impersonating the tax registry. No real surprise, given deadlines for various tax-related forms to be submitted fall around this time of year.
Users receive an email informing them about 'tax discrepancies' stating immediate action is required. They are informed they can either go directly to the 'Agenzia Entrate' (Internal Revenue Service) website, or access the file attached to the email which has a .URL file extension and is password protected, with the password made available within the body of the email.
The links within the email are designed to access a directory on a web server using the SMB protocol which will subsequently connect to another URL to download a ZIP file containing a malicious CPL (Control Panel) file, which in turn will trigger the start of the infection chain.
Emails subjects include the following:
Comitato di osservazione dell'anagrafe tributaria
Commissione di monitoraggio del registro tributario
Gruppo di controllo del registro tributario
Comitato di monitoraggio dell'anagrafe tributaria
Comitato per l'osservanza dell'anagrafe tributaria
Organismo di supervisione sull'anagrafe tributaria
Symantec has 0-day protection in place, identified by the following:
File-based
Trojan.Downloader
Trojan Horse
Trojan.Mallnk
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers.
Web-based
Observed domains/IPs are covered under security categories in all WebPulse enabled products.
Click here to learn more about Symantec's Endpoint Security Service.
Click here to learn more about Symantec's Cloud Email Security Service.
Click here to learn more about Symantec's cloud-based Web Security Service (WebPulse). |
Protection Highlight: VB6 Threats Still Active In 2024
Per Wikipedia, Classic Visual Basic (commonly referred to as "VB"), first released in 1991, is a third-generation programming language based on the BASIC programming language (released way back in 1963!) and an integrated development environment (IDE) from Microsoft for Windows. The final release was Version 6 (VB6) in 1998. On April 8, 2008, Microsoft stopped supporting the VB6 IDE, releasing in its place VB.NET, designed for Microsoft's new . NET Framework which addresses many of the limitations of the classic model.
Even though more than 25 years have passed, VB6 executables can still run on the latest Windows versions due to Microsoft ensuring compatibility for VB6 applications on supported Windows operating systems. We still see many VB6-based systems running in factories, hospitals and businesses thanks to that compatibility. On the flip side however, attackers can also still utilize VB6 to create new viruses, probably in the hope that it will be harder for security researchers to analyze and detect these executable files crafted using old school tools.
The Vilsel worm is one such VB6 malware that periodically appears in the threat landscape. Since it was first reported by security researchers in late 2015 Vilsel has been modified several times and we still see its presence globally in 2024, part of the overall VB malware ecosphere.
Vilsel is built by utilizing PCode, or Pseudo-code, within the aging Visual Basic compiler build options. PCode is not a native CPU code but VB6-specific intermediate code that is executed by Microsoft's compatibility program step by step. Thus many of the commonly used modern analysis tools do not work for investigating these files - a VB decompiler or similar tool being required to analyze them. Some variants of Vilsel further encapsulate a VB6 executable inside a native code obfuscator, causing some less robust anti-VB6 protections to malfunction.
One specific variant of Vilsel we analyzed works like this:
Creates a copy of itself, appending some extra bytes at end of the code
Copies itself to an existing folder, repeating this for all folders in the system
Uses one of the following file names: backup.exe, System Restore.exe, update.exe, data.exe
Interestingly, System Restore.exe, update.exe and data.exe are chosen at a probability of 1/30 respectively, hence backup.exe is chosen at a probablity of 27/30 or 90%, meaning this will most likely be the name the file ends up using to hide itself among similarly named, legitimate system files.
Symantec protects you from this threat, identified by the following:
File-based
W32.Vilsel!gen1
W32.Vilsel!gen2
Click here to learn more about Symantec's Endpoint Security Service. |
Protection Highlight: What did SEP IPS do for you last year (2023)?
Symantec's IPS is a best-in-class deep packet inspection engine, protecting hundreds of millions of endpoints (desktops and servers) including Fortune 500's and consumers.
We've posted about the virtues of using IPS in your defense posture many times - from blocking tech support scams, to putting the brakes on endless Log4j attacks, to phishing, SMB attacks, CMS vulnerabilities, the highly effective IPS browser extension, and the list goes on. To quote a comment from that last post, "if your security setup doesn't include Intrusion Prevention you are potentially exposing your organization to a significant loss of threat protection".
Let's take a quick look back at the year past to see how IPS helped protect our customers.
During 2023, SEP's network protection engine (IPS) blocked over 4.7 billion threats on Symantec protected endpoints. 93% of these attacks were blocked at the pre-infection stage.
Attempts to exploit Web Server Vulnerabilities: 38%
Attempts to exploit Windows OS Vulnerabilities: 20%
Malware C&C attempts: 17%
Attempts to perform Web Payload Uploads: 6%
Attempts to redirect users to attacker-controlled websites: 6%
Attempts to exploit Server Vulnerabilities: 3%
Coin mining attempts: 3%
Attempts to exploit Application Vulnerabilities: 2%
Pre-ransomware attacks: 2%
Attempts to exploit CMS Vulnerabilities: 1%
Other: 2%
Customers are advised to enable IPS on Desktops and Servers, for best protection. Click here for instructions on enabling IPS.
Click here to learn more about managing Symantec's Endpoint Security Intrusion Prevention System (IPS). |
Protection Highlight: What did WebPulse do for you last month?
Symantec WebPulse is a best-in-class Web Security engine, protecting hundreds of millions of customer devices including Fortune 500's and consumers.
Part of the Symantec Global Intelligence Network (GIN), WebPulse is a cloud-based infrastructure specifically designed to harness the power of user-driven behavior and to translate user input into global web intelligence and web threat intelligence.
WebPulse uses multiple technologies to analyze this input to deliver the fastest and most accurate web categorization and risk levels of any vendor. Within the WebPulse framework, each incoming URL request is processed by many different threat analysis methods, both automated and manual.
In the last 30 days, WebPulse blocked a total of 777M attacks across a range of different security categories.
* Note these numbers only cover the WP cloud service and don't include on-prem devices
456M in the Suspicious category
173M in the Malicious Sources/Malnets category
49M in the Phishing category
25M in the Compromised Sites category
22M in the Malicious Outbound Data/Botnets category
21M in the Scam/Questionable Legality category
17M in the Spam category
14M in the Potentially Unwanted Software category
Customers are advised to enable WebPulse real-time categorization, for best protection. Click here for instructions on configuring WebPulse.
Click here to learn more about WebPulse, Symantec's cloud-based Web Security engine. |
Protection Highlight: WikiLoader Returns
WikiLoader was first discovered back in 2022, a downloader that drops a separate malware payload onto the target machine. It contains several evasion techniques along with custom code and has been associated with two primary threat actor groups, TA544 and TA551.
It was named WikiLoader as the original variant made an outbound request to the Wikipedia website, most likely to check that it had an Internet connection and wasn't being run inside a virtual machine or isolated environment which may indicate it's being analyzed by a security researcher or some type of automated analysis system.
The typical infection chain for WikiLoader starts with an email containing a PDF attachment. Within the PDF file, there will be a link that if accessed will download a compressed JavaScript file which will in turn download the final payload. It's worth noting the extraordinary lengths the author went to attempting to obfuscate the script. Of the original file of over 7 MB and 4383 lines of code, approximately 99.8% is obfuscation (i.e. junk) aimed at trying to hide the remaining tiny piece of core executable downloading logic. Given the main purpose of WikiLoader - malware for rent - that final payload could be anything, but to date tends to be banking malware such as Ursnif or some type of information stealing malware.
Email subject lines observed:
Invoice from [company name] S7354534 , 01070875, Customer ref: A0627 15.12.23, US
Invoice from [company name] S8786164 , 04725130, Customer ref: A0703 15.12.23, US
Example email:
Symantec protects you from this threat, identified by the following:
File-based
Scr.DLHeur!gen8
Scr.Malcode!gen
Web.Reputation.1
WS.SecurityRisk.4
Network-based
[33403] Audit: System Process Accessing discordapp.com
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers.
Web-based
Observed domains/IPs are covered under security categories.
Click here to learn more about Symantec's Endpoint Security Service.
Click here to learn more about managing Symantec's Endpoint Security Intrusion Prevention System.
Click here to learn more about Symantec's Cloud Email Security Service.
Click here to learn more about Symantec's cloud-based Web Security Engine (WebPulse). |
Protection Highlight: XMRig Miner
XMRig is a popular open-source CPU-based cryptocurrency mining application that is designed to mine the Monero (XMR) cryptocurrency. While XMRig is a legitimate tool, it is often used by malware authors to mine Monero using (i.e. stealing) the processing power of compromised systems without the user's consent. This practice is called cryptojacking. From this point on when we mention XMRig, we are referring to the cryptojacking instance.
XMRig, prevalent though it is, is only one of many unwanted crypto miners, some blocked first by our AV (file analysis) engine technology and others caught by our IPS (network traffic analysis) technology, depending on the particular infection vector. The use of "first" here because most would be blocked by both protection technologies, but once a threat has been neutralized it generally won't trigger subsequent protection technologies and therefore the first to block it gets the credit. That said, some components of a threat might be blocked by one protection type and other components of the same threat blocked by a different tech. This really underscores the importance of "layered security".
The below chart shows the AV/IPS breakdown over the past month.
XMRig - included here - is just one of many unwanted crypto miners Symantec blocks.
XMRig will likely use somewhere between 70% and 80% of a compromised computer's central processing unit (CPU) and/or graphics processing unit (GPU) power. When XMRig is running, a user may notice that their computer is running slower than usual and games or applications are intermittently freezing. Depending on what else the computer is being used for at the time, it may also get quite hot, potentially causing damage to the hardware components. Prolonged use will also consume significantly more power than normal which could increase electricity bills.
There are a few reasons why XMRig is a popular choice for malware authors:
* XMRig is relatively easy to use and configure, making it accessible even to novice cybercriminals.
* XMRig is open-source, meaning that its source code is freely available for anyone to modify and distribute. Malware authors can easily modify the code in an attempt to evade detection and increase their profits.
* Monero is a popular cryptocurrency for mining because it is designed to be mined with CPU power, making it easier to mine on a large scale using a botnet of compromised systems.
* XMRig is efficient and can mine Monero at a relatively high rate without consuming excessive amounts of system resources, making it ideal for cryptojacking.
The XMRig miner spreads via a wide variety of methods, including Phishing and other types of email spam, malvertising, malicious droppers, vulnerabilities, cracked software applications, Potentially Unwanted Applications (PUA), Drive-by-download, etc.
Symantec has had long-standing detection of XMRig and strong 0-day protections are in place:
File-based
CL.XMRig!gen1
Miner.XMRig
Miner.XMRig!gen*
OSX.Miner.XMRig!g1
Behavior-based
SONAR.Suspdrop!g61
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers.
Machine learning-based
Heur.AdvML.*
* This indicates multiple similarly named detections exist such as Heur.AdvML.B, Heur.AdvML.C, etc.
To learn more about Symantec's Endpoint Security Service, click here.
To learn how Symantec behavioral security technologies provide protection, click here.
To learn more about Symantec's Cloud Email Security Service, click here.
To learn more about how Symantec Endpoint Protection uses advanced machine learning, click here. |
Proton Ransomware
Over the past few months, yet another Ransomware actor has been observed in the already crowded threat landscape. Known as Proton, according to the ransom note the threat actor doesn't appear to employ double-extortion tactics. When machines are successfully encrypted, files will find themselves appended with a .proton or .Kigatsu extension. No specific amount is specified in the ransom note. As of today, it is simply an FAQ that provides victims with an ID and includes the actor's Telegram and email contacts.
Symantec protects you from this threat, identified by the following:
Behavior-based
AGR.Terminate!g2
File-based
Trojan.Gen.MBT
Machine Learning-based
Heur.AdvML.B |
PS99, hopes for free diamonds leads to ransomware
Pet Simulator 99 is the fourth installment in the Pet Simulator series developed by BIG Games on the Roblox platform. The game revolves around farming coins and diamonds to unlock powerful pets, with coins used for hatching. Launched on December 1st, 2023, many players attempt to use game hacks for auto-farming diamonds, unwittingly exposing themselves to risks.
In a recent incident, Symantec observed a drive-by-download campaign wherein an actor disguised Chaos ransomware as a fake Auto Diamond farming tool. If successfully executed, this malicious tool will encrypt files, prompting victims to contact the attackers via email or Discord.
Symantec protects you from this threat, identified by the following:
File-based
Trojan Horse
Trojan.Gen.MBT
WS.Malware.1
Machine Learning-based
Heur.AdvML.B
Heur.AdvML.A!300
Heur.AdvML.B!100
Heur.AdvML.B!200 |
PseudoManuscrypt spyware distributed in widespread campaigns
According to a recently released report, several campaigns distributing the PsedoManuscrypt malware have been observed across the year 2021. While PsedoManuscrypt bears some similarities to the Manuscrypt malware leveraged in the past by the Lazarus APT group, it does not show any indications of particular targeting. Attackers behind this malware have been leveraging MaaS platforms to distribute the loader hidden within pirated software installer archives. PseudoManuscrypt has spyware functionalities including keylogging, screenshot grabbing or clipboard data theft among others.
Symantec protects you from this threat, identified by the following:
File-based
Trojan Horse
Trojan.Gen.2
Trojan.Gen.MBT
Machine Learning-based
Heur.AdvML.B
Heur.AdvML.C
Network-based
Web Attack: Webpulse Bad Reputation Domain Request
Web-based
Observed domains/IPs are covered under security categories |
Public Alert System exploited to deploy SpyNote Android malware in Italy
'IT-alert' is an SMS-based Public Alert System, an Italian Government-owned service designed to send out warnings about an imminent or ongoing natural disaster to the general public. A report shows that attackers leveraged the service and sent out a fake warning about a possible volcanic eruption. Users were urged to install an application to keep them updated about the activities, in actuality unwittingly installing a SpyNote malware capable of stealing banking, cryptocurrency and social media credentials. Other capabilities include camera recording, GPS & network location tracking, and recording phone calls.
Symantec protects you from this threat, identified by the following:
File-based
WS.Malware.1
Mobile-based
Android.Reputation.2
Web-based
Observed domains/IPs are covered under security categories in all WebPulse enabled products |
Publishing application services abused to spread phishing runs
Symantec has already observed a plethora of legitimate web applications been abused by threat actors to spread phishing or malicious spam. Adding to this list, Symantec observed a publishing web application – Lucidpress - being abused to host phishing content. As shown in Figure 1, emails posing as fax notification messages are sent to recipients asking to review and compete fax information. The email content is kept short and the recipient is provided with a hyperlinked text to click for review.
Figure 1: Phishing Email masquerading as a Fax Notification with abuse of lucidpress.com URL in hyperlinked text
As shown in Figure 2, upon clicking the hyperlinked text, it redirects to the Lucidpress’s online document viewer. This is where the phishing document is hosted and contains a phishing URL embedded within the file content.
Figure 2: Phishing .pdf file hosted on Lucidpress’s online document viewer.
On clicking the hyperlinked text in the file, a credential stealing webpage is served to the user as shown in Figure 3.
Figure 3: Phishing page spoofing reputed brands / services
Protection
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers. |
Pupy RAT continues to be used in attacks against Linux systems
Pupy RAT continues to be leveraged in attacks conducted by miscellaneous threat operators. The malware has various functionalities including upload/download of files, remote command execution, information theft, keylogging and screenshot capture among others. While Pupy RAT is known to target both Windows and Linux systems, recently reported campaigns have seen usage of the Linux variant of this malware against targets in Asia.
Symantec protects you from this threat, identified by the following:
File-based
Packed.Vmpbad!gen38
Trojan.Gen.MBT
Trojan.Gen.NPE
WS.Malware.1
Machine Learning-based
Heur.AdvML.B!100
Heur.AdvML.B!200
Heur.AdvML.C
Web-based
Observed domains/IPs are covered under security categories in all WebPulse enabled products |
Pure Clipper malware targets Italian-speaking users
A new campaign delivering Pure Clipper malware to Italian-speaking users has been observed in the wild. The malware is attributed to a threat group known as Alibaba2044, which has also been targeting Italy in older attacks leveraging PureLogs malware. In this latest campaign, Pure Clipper has been spread via a phishing website impersonating the official Tor Project webpage. The malware has the functionality to swap cryptocurrency wallet addresses copied by the users with those belonging to the attackers.
Symantec protects you from this threat, identified by the following:
Behavior-based
AGR.Terminate!g2
SONAR.Powershell!gen4
File-based
Infostealer.Limitail
Scr.Malcode!gdn14
Scr.Malcode!gdn32
Scr.Malcode!gen
Trojan.Gen.MBT
Machine Learning-based
Heur.AdvML.A!400
Heur.AdvML.B!100
Heur.AdvML.B!200
Web-based
Observed domains/IPs are covered under security categories in all WebPulse enabled products |
PureCrypter Loader
PureCrypter is a loader that provides various groups and individuals with the capability to add another level of evasion in their attack chain in order to deploy (via injection) their payloads. Since it was first advertised and sold on underground markets around March 2021, it quickly gained in prevalence and has been observed deploying various remote access trojans and infostealers.
Symantec protects you from this threat, identified by the following:
File-based
Trojan.Gen.2
Trojan.Gen.MBT
Trojan Horse |
PureLand Stealer for macOS
PureLand is an infostealer variant targeting the macOS platform. The malware was initially discovered in early 2023 and seen being distributed alongside of Redline Stealer variant targeting Windows machines. PureLand has the functionality to collect host information, cookies, extract data from various browser wallet extensions or crypto-wallet applications, among others.
Symantec protects you from this threat, identified by the following:
File-based
Trojan Horse
OSX.Trojan.Gen
OSX.Trojan.Gen.2
WS.Malware.1 |
PureLogs stealer observed in malspam campaign
There is a wide variety of malware-as-a-service available in the current threat landscape. PureLogs stealer, a type of infostealer, is one such example that has recently been discovered through careful analysis of its activities. It has been reported that this malware is being distributed through malicious emails in Italy.
The creators of PureLogs stealer are known as "Pure Coder." According to their advertisements, they also offer other types of malware services, including a crypter, miner, loader, and remote access trojan. These types of malware-as-a-service are becoming increasingly common, and it is important for individuals and organizations to be aware of the various ways in which they can be transmitted and the potential risks they pose.
Symantec protects you from this threat, identified by the following:
Behavior-based
SONAR.Mshta!g1
SONAR.PsDownloader!g1
SONAR.SuspScript!g5
File-based
Trojan.Gen.MBT
Trojan.Gen.NPE
Suspicious: Content
Machine Learning-based
Heur.AdvML.B
Network-based
System Infected: Activity - Bad Application Reputation Application 7
System Infected: Trojan.Backdoor Activity 634
Web-based
Observed domains/IPs are covered under security categories |
Pure Malware Family
Pure is a malware family active on the threat landscape since at least 2021 and attributed to a developer known as PureCoder. PureCrypter and PureLogs infostealer are the better known malware variants belonging to this family. PureCrypter assists the attackers with encryption and obfuscation capabilities which add additional level of evasion in the conducted attacks. PureLogs on the other hand is an infostealer variant focused on data ex-filtration including machine information, browser data, cryptowallets, and others. The latest arrival to the Pure family is PureMiner - a cryptomining malware used to deploy XMRig miners on the infected endpoints.
Symantec protects you from this threat, identified by the following:
Behavior-based
AGR.Terminate!g5
AGR.Terminate!g7
SONAR.SuspBeh!gen667
SONAR.SuspLoad!g59
File-based
Trojan.Gen.2
Trojan.Gen.MBT
WS.Malware.1
Machine-Learning based
Heur.AdvML.A!300
Heur.AdvML.A!400
Heur.AdvML.A!500
Heur.AdvML.B!200
Heur.AdvML.B!100
Heur.AdvML.C
Web-based
Observed domains/IPs are covered under security categories in all WebPulse enabled products. |
PurpleFox (aka DirtyMoe) malware distributed to users in Ukraine
PurpleFox malware known also as DirtyMoe has been reported to be actively distributed to users in Ukraine. DirtyMoe has modular architecture and comes with a rootkit component. The malware has backdoor capabilities, can be used for remote access, download of additional arbitrary payloads or even for execution of DDoS attacks. The infection chain involves malicious .MSI installers in the initial stages and malware self-propagation mechanisms through brute-force attempts and vulnerability exploitation in later attack phases.
Symantec protects you from this threat, identified by the following:
Adaptive-based
ACM.Ps-Rd32!g1
Behavior-based
SONAR.TCP!gen1
File-based
Hacktool
Hacktool.Rootkit
Infostealer
Trojan Horse
Trojan.Gen.MBT
Trojan.Kewgad
WS.Malware.1
WS.Malware.2
Machine Learning-based
Heur.AdvML.A!300
Heur.AdvML.B!100
Heur.AdvML.B!200
Web-based
Observed domains/IPs are covered under security categories in all WebPulse enabled products |
Purple Fox Exploit Kit re-charges with new exploit codes
Exploit Kits are a common way by which attackers can distribute their malicious code to unsuspecting targets. While the use of exploit kits has been generally in decline over the course of the last few years, Purple Fox is one of the relatively newer kits that has already shown some spikes this year. It has also been updated a few times with the inclusion of additional exploits. Those most recently seen being used by Purple Fox are:
CVE-2020-1054 (recently added)
CVE-2019-0808 (recently added)
CVE-2020-0674
CVE-2019-1458
Symantec protections against Purple Fox Exploit Kit activities:
File-based
Exp.CVE-2020-0674
Exp.CVE-2019-0808
Exp.CVE-2019-1458
Hacktool
ISB.Heuristic!gen5
Trojan.Gen.MBT
Trojan.Gen.NPE
Trojan Horse
Network-based
Web Attack: Microsoft Internet Explorer CVE-2020-0674
Web Attack: Suspicious Executable Image Download
Website-based
This attack is proactively blocked by Symantec Web Isolation technology |
Purple Fox exploits weak SMB passwords
As Security Response reported in a previous protection bulletin, Purple Fox malware recently added worm capability to its arsenal. The attackers did not stop there - its armory has been further enhanced with SMB password brute force for propagation. Weak passwords such as "password" and "1234567890" must be updated in order to avoid becoming a victim and providing a stepping stone for other attacks.
Symantec protects you from this threat, identified by the following:
File-based
Trojan.Gen.MBT
WS.Malware.1
WS.Malware.2 |
Purple Fox now worms
Purple Fox (related to Purple Fox Exploit Kit) has been around for a while now. A downloader with backdoor and rootkit capabilities, it has recently been reported to have added worm capability.
It was known that actors behind this threat have been using Exploit kits and Email as vectors but recently with this new worm capability it has introduced a third vector of infection. They can now scan Internet-facing Windows machines and potentially obtain entry into the network via SMB password brute force.
Symantec protects you from this threat, identified by the following:
File-based
Trojan.Gen.MBT
WS.Malware.1
WS.Malware.2 |
Purple Fox Rootkit Infection
A recent malware campaign was discovered running a malicious installer compiled using AutoIt and named "Telegram Desktop.exe". If triggered, two files will be dropped, a legitimate Telegram installer file 'Telegram.exe', and a malicious downloader 'TextInputh.exe'. The AutoIt script will help launch TextInputh.exe and the next stage of the attack will commence. Connection with the hardcoded C2 server will be established to download two additional files, one being an archive (RAR) file. This archive file contains the final payload which includes the Purple Fox Rootkit.
If successfully deployed, the Purple Fox Rootkit will disable the User Account Control (UAC). UAC is a vital security feature to help prevent unauthorized accesses and helps mitigate the impact of malware.
Symantec protects you from this threat, identified by the following:
File-based
Hacktool.Rootkit
Trojan.Gen.MBT
Trojan.Gen.2
Trojan Horse
W97M.Downloader
Web-based
Observed domains/IPs are covered under security categories |
PurpleWave infostealer found in cybercrime forums
Purplewave is a recently uncovered infostealer that is being sold on Russian cybercrime forums. This infostealer can be purchased with either lifetime support or support for two updates. The cost is roughly between $50-70.
The infostealer contains the standard functions expected of this type of malware, including, but not limited to:
Stealing system files/information
Stealing cryptocurrency wallet data
Stealing data from Chromium and Mozilla browsers
Screen captures
Installation/Execution of additional payloads
Symantec protections
File based
Heur.AdvML.B
Trojan Horse
Trojan.Dapato
Trojan.Gen.2
Trojan.Gen.MBT
Network based
System Infected: Trojan.Backdoor Activity 495 |
Puspa (aka Chaos 500) Ransomware
A ransomware called Puspa (also known as Chaos 500) is not new to the threat landscape but is only observed sporadically and fortunately, is not as disruptive as many of its more notorious peers. According to their most recent activity, the actor behind this threat infects single machines and requests a ransom of 500 US dollars in Bitcoin.
Symantec protects you from this threat, identified by the following:
Behavior-based
SONAR.SuspDrop!gen1
File-based
Ransom.Sorry
Machine Learning-based
Heur.AdvML.B |
Putting the cuffs on RedLine Stealer
RedLine Stealer, a relatively new infostealer spotted earlier this year in the email threat landscape continues its nefarious activity. While Symantec has observed it being delivered via drive-by-download, as of today email is one of the primary vectors by which it spreads.
Below is a list of email subjects we have seen in September and early October associated with multiple RedLine spam campaigns:
REQUEST FOR QUOTATION
INQUIRY Urgent need here with our clients now.
Purchase Order Details JJFL8092020 for the Month Of Sep-2020
Purchase Order MT (Quotation Request)
Remittance for shipment - GMP 0004224 20MT
Symantec has the following coverage in place:
File-based
Trojan Horse
Trojan.Gen.NPE
WS.Malware.1
Heur.AdvML.B
Heur.AdvML.C
Behavior-based
SONAR.SuspLaunch!g12
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers.
Website-based
This attack is proactively blocked by Symantec Web Isolation technology |
PYbot DDoS malware
PYbot is a Python-based DDoS malware targeting Windows environments. The malware is distributed under the disguise of a Discord Nitro token generator software called Nitro Generator. PYbot supports various types of DDoS attacks including TCP Flood, TCP SYN Flood, UPD Flood, VSE (Valve Source Engine) Flood and HTTP GET Request Flood.
Symantec protects you from this threat, identified by the following:
Behavior-based
SONAR.TCP!gen1
File-based
Trojan Horse
Trojan.Gen.MBT
WS.Malware.1
Machine Learning-based
Heur.AdvML.B
Web-based
Observed domains/IPs are covered under security categories |
PyMICROPSIA targets the Middle Eastern region
Symantec Security Response is aware of a new infostealer that was used for cyber attacks in the Middle East. The malware is dubbed PyMICROPSIA because of it being written in Python as well as its similarity to another malware "MICROPSIA" which the same threat actor weaponized. PyMICROPSIA is not only capable of collecting information and Outlook .ost files from the infected machine but can also download and execute additional payloads.
Symantec provides the following coverage against PyMICROPSIA:
File-based
Infostealer
Trojan Horse
Trojan.Gen.2
Trojan.Gen.MBT |
Pysa aka Mespinoza among the most active ransomware variants in November
According to a recently released report, Pysa aka Mespinoza ransomware has been identified as one of the most active ransomware variants targeting organizations in November. Attackers behind Pysa are known to employ the Double Extortion tactics - namely exfiltrating victim's data before encryption and threatening to release it publicly if the ransom demands are not met.
Symantec protects you from this threat, identified by the following:
File-based
Downloader
ISB.RansomMespinoza!g1
Ransom.Gen
Ransom.Mespinoza
Behavior-based
CL.RansomMespinoza!g1
CL.RansomMespinoza!g2
SONAR.Ransom!gen57
SONAR.SuspWrite!g4
Machine Learning-based
Heur.AdvML.B
Heur.AdvML.C |
PYSA ransomware attack
Symantec Security Response is aware of a recent alert issued by the FBI warning of an increase in PYSA ransomware infections at educational institutions in the United States and United Kingdom. PYSA ransomware, also known as Mespinoza, has been around since at least March 2020 and adds a .pysa file extension to the files it encrypts. The ransomware also steals confidential information prior to the encryption, which is then used to further entice the victim into paying a ransom.
Symantec provides the following protection against PYSA ransomware:
File-based:
Ransom.Gen
Ransom.Hermes!gen2
Downloader |
Pysa ransomware threat group - now also targeting Linux
It has been reported that the threat group behind the Pysa ransomware has also been targeting the Linux environments. A new linux variant of the Chachi Golang-based malware leveraged by the group has emerged in the wild. Chachi backdoor has been a part of Pysa's arsenal already in the past and is used for network enumeration, lateral movement across networks and data ex-filtration among others.
Symantec protects you from this threat, identified by the following:
File-based
Trojan.Gen.NPE
Network-based
Web Attack: Webpulse Bad Reputation Domain Request
Web-based
Observed domains/IPs are covered under security categories |
PYSA using new GoLang RAT ChaChi
A more refined variation of ChaChi has been seen since the start of 2020. This malware is now a custom RAT that uses a new programming language "Go". PYSA (aka Mespinoza) threat operators use the new ChaChi to perform typical RAT activities such as backdoor creation, defensive evasion of essential services, data exfiltration/credential dumping, network enumeration, service installation, and lateral movement across networks.
This trojan has been part of PYSA's arsenal to attack victims globally across a variety of industries for some time, with most recent targets of focus being healthcare and education organizations.
Symantec protects you from this threat, identified by the following:
File-based
Backdoor.Trojan
Heur.AdvML.B
Trojan Horse
Trojan.Gen.2
Trojan.Gen.MBT
WS.Malware.1
WS.Malware.2
Web-based
Observed IPs/Domains are covered under security categories
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers. |
Pyschedelic ransomware
Another run-of-the-mill ransomware known as Pyschedelic has been seen distributed in the wild. While the malware name seems to be misspelled, it has been provided by the threat actors as such in the ransom notes. The ransomware encrypts files on a victim’s machine and requests the victims to contact the attackers via email. Pyschedelic ransomware leverages Windows “certutil -encode” command to encode files.
Symantec protects you from this threat, identified by the following:
File-based
Scr.Malcode!gdn32
Trojan.Gen.MBT
WS.Malware.1
Machine Learning-based
Heur.AdvML.B |
Python based EvilAnt Ransomware
Evil Ant Ransomware is a Python based malware compiled using PyInstaller. In addition to performing the basic Ransomware activities of file encryption and displaying a ransom note, Evil Ant also has anti-analysis (won't run in a Virtual Machine) and anti-detection (disables Windows Defender) capabilities. Of note is that this ransomware is rather large in size, reporting in at around 27MB.
Symantec protects you from this threat, identified by the following:
Adaptive-based
ACM.Untrst-RunSys!g1
File-based
Ransom.Gen
Trojan.Gen.MBT
Machine learning-based
Heur.AdvML.A
Heur.AdvML.C |
Qakbot lives on
Qakbot is a well-known and relatively sophisticated banking malware that has plagued the threat landscape for years. The authors behind this malware have been using email and exploit kits as infection vectors, and as of today, Symantec continues to observe campaigns delivering this malware.
The following protections are in place:
File-based
W32.Qakbot
Trojan.Horse
W97M.Downloader
Packed.Generic.534
SAPE.Qakbot.1
Heur.AdvML.B
Heur.AdvML.C
WS.Malware.1
WS.Malware.2
Network-based
System Infected: W32 Qakbot Activity*
Behavior-based
SONAR.Qakbot!*
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers.
Website-based
This attack is proactively blocked by Symantec Web Isolation technology
* This indicates multiple detections exist such as SONAR.Qakbot!g1, SONAR.Qakbot!g2, etc... |
Qakbot rides the US Election wave
Symantec Security Response has become aware of a Qakbot spam campaign attempting to leverage the US Presidential Election. The malware is delivered via an email with a US Election themed attachment named "ElectionInterference_[8 to 9 digits].zip". If the infection is successful Qakbot will reach out to its command and control server for further instructions.
Symantec detects the malicious attachments and Qakbot as the following:
File-based
Trojan.Mdropper
W97M.Downloader
Backdoor.Trojan
Heur.AdvML.C
Network-based
22819 (Web Attack: Suspicious Executable Image Download)
29844 (System Infected: W97M.Downloader Activity 27)
Email-based
Coverage is in place for Symantec's email security products and Email Threat Isolation (ETI) technology provides an extra layer of protection for our customers. |
qBit Stealer
qBit Stealer is a malware developed in Golang by the qBit ransomware group. The malware source code was recently published on underground forums for reuse. This malware incorporates functionality to collect sensitive user data and system information, which is then exfiltrated to the MEGA file-sharing service. Additionally, it employs anti-debugging and anti-virtualization techniques to enhance evasion.
Symantec protects you from this threat, identified by the following:
File-based
Trojan.Gen.MBT
WS.Malware.1
Policy-based
Symantec Data Center Security (DCS) provides Multi-layered protection from this malware
Suspicious Process Execution: Prevention policies prevent malware from being dropped or executed on the system
Network Control: Prevention policies block outgoing connections to public internet ( Mega[.]nz) in this case |
Qilin Linux ransomware variant targets VMware ESXi
Qilin ransomware, also known as Agenda, was initially discovered back last year and been targeting Windows instances. A new Linux based variant of this malware has been observed recently in the wild and reported to be targeting VMware ESXi instances. The malware incorporates configuration settings allowing for selective encryption of specified file types as well as setting folders excluded from encryption. Qilin has the functionalities to terminate selected system processes and running VMs as well options for additional customization of the VM snapshot encryption process.
Symantec protects you from this threat, identified by the following:
File-based
Ransom.Qilin
WS.SecurityRisk.4 |
Qilin ransomware, another ransomware moving to Rust
Qilin Ransomware, also known as Agenda, was previously found written in Go language, however a new version was found written in Rust. We have seen several malware and ransomware making the change to Rust for easier adjustment to different OS platforms and in an attempt to make more difficult its analysis. The Rust version seems to lack some of the features previously seen in the Go version, which may mean that they are still developing and adjusting this new version or that they are trying to simplify or modify its operations.
Symantec protects you from this threat, identified by the following:
File-based
Ransom.Qilin
Ransom.Qilin!g1
Trojan.Gen.MBT
WS.Malware.1
Machine Learning-based
Heur.AdvML.C |
Qilin Ransomware
Qilin Ransomware (also known as Agenda) which employs double-extortion tactics, has recently been spotted in the threat landscape. Over the past few weeks it has been increasing its activity, targeting various organizations around the globe with a focus on Latin America. The ransom note contains information on what type of data has been encrypted/stolen and credentials to connect to the actor's Onion domain.
Symantec protects you from this threat, identified by the following:
File-based
Trojan.Gen.MBT
Ws.Reputation.1
Machine Learning-based
Heur.AdvML.C |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.