wildercb/PrivacyTextBooks1.2 · Datasets at Hugging Face

text stringlengths 0 118
because they are no longer updated. Yet they remain connected to networks,
making it easier for criminals to take advantage of unpatched
vulnerabilities.These unserviced software programs and devices are like
radioactive waste; they can’t just be left behind or thrown away. When
technology companies deprecate software, they must have a meaningful
plan to transition people away from using it. Merely saying “stop” isn’t
good enough. Companies need to ensure that people don’t lose their
investments in the software, such as source files or custom software.
The law could better facilitate such reform of the deprecation process if
it held technology companies more accountable.
Sending Sensible Signals
Design doesn’t just facilitate or hinder tasks. It also delivers information. It
communicates by sending a signal to the user of a technology.65 Through
signals, design helps define our relationships with other people and shapes
our risk calculus when interacting online. These signals affect our
expectations about how things work and the context within which we are
acting.
Data security law focuses a lot on safeguards like encryption but hardly
at all on the signals that technologies send to the people using them, like
whom a communication is coming from or how a feature works or should
be used. Yet, these signals are what encourage people to click certain links,
share personal data, or enable certain features. Designers must engineer
their systems so that the signals are understandable and that they encourage
secure rather than insecure behavior.
The wrong design can send misleading signals. People will ignore
signals if too many are sent or if not enough are sent. Poor signals can be a
major vulnerability. Fraudsters exploit confusing, inconsistent, and
ambiguous signals to trick people. In essence, poor signals make people
more gullible.
MISLEADING SIGNALS
When people use devices, there is a ton of information thrown their way
about the security risks of various activities. Unfortunately, this information
is often vague or misleading, making it difficult for people to get a clear
sense of what they should do or how vulnerable they really are.
One example is the ubiquitous padlock icon people see when they use
their browser, adjust their privacy settings on social media, and enter
authentication credentials.66 The padlock is an icon of the physical
manifestation of security—only those with the key get access to whatever it
is protecting. But what does it mean in specific contexts online? It could
mean almost anything, from the deployment of specific encryption and
authentication protocols to a general warm and fuzzy sense of “security”
similar to the comically vague statements in privacy policies that a
“company takes your privacy and security seriously.” If nothing else,
padlock icons are invitations to garner consumer trust, enticing people not
to worry about providing more personal information.67 The padlock icon
isn’t necessarily good or bad, but it is vague and companies often use it to
promise more than they deliver. Because this icon isn’t regulated, the trust
that it conveys is often false and prone to abuse.
Policymakers can play an important role in encouraging more useful
signals and discouraging vague and misleading ones. At a minimum,
regulators should provide guidance and facilitate industry coherence in
security signals for users. Regulators can also do a lot more, such as
applying federal and state laws against deceptive trade practices to
companies that use false or misleading signals. In several cases, the FTC
has alleged that technology companies using phrases like “easy to secure”
and “advanced network security” were being deceptive because their
products and services were insecure.68 Icons that invoke the concept of
security are conveying similar messages even though words aren’t being
used.
The term “security” itself also functions as a signal. Many promises
made by companies about security are as vague and empty as the padlock
icon. In other areas, the FTC has created rules to limit when companies can
use certain terms that might mislead consumers. For example, the FTC has
limited the extent to which companies can use the word “free” to describe
certain products and services.69 The use of the word “security” and icons to
represent security should be similarly scrutinized.
POORLY TIMED, TOO MANY, OR NOT ENOUGH SIGNALS
Signals not only fail when they are inconsistent, vague, or misleading.
Signals also fail based on the timing and frequency of their use. For
example, researchers at University College London conducted a study to
ascertain the effectiveness of security warnings. They found that too many
warnings desensitized people to the risks. The researchers concluded that
“security warnings in their current forms are largely ineffective, and will
remain so, unless the number of false positives can be reduced.”70 In
another study, a different group of researchers concluded that “the status
quo of warning messages appearing haphazardly—while people are typing,
watching a video, uploading files, etc.—results in up to 90 percent of users
disregarding them.”71 Researchers at Carnegie Mellon’s CyLab developed
Warning Design Guidelines, which recommended that warnings be clear,
concise, and accurate: “If too long, overly technical, inaccurate, or
ambiguous, a warning will simply be discarded and its purpose will be
lost.”72 These studies and others have repeatedly shown that security
warnings are like the porridge for Goldilocks—they have to be just right.
Too many warnings will work poorly, as will too few. Warnings at the
wrong place and time will work poorly.
Warnings should reflect the gravity of the risk. Using similar types of
warnings for low security risk and for high security risks fosters confusion;
people might begin to assume that all the warnings are for low risks and can
be ignored. Warnings must be implemented in ways to avoid being treated
like the boy who cried wolf. Moreover, the CyLab Guidelines suggest that
warnings “follow a common visual layout” because it “can be recognized
faster.”
Instead of better calibrating warnings to the risk, the opposite trend is
occurring. Some companies are now slapping warnings on all external
emails, with warnings like: EXTERNAL EMAIL—SPAM RISK. As these
warnings adorn so many harmless messages, people will become
desensitized to them.
Another problem is with spam filters. When going through the junk mail
folder to look for any legitimate emails that have been snagged, there is
often nothing to help distinguish the danger of the emails in the folder. All

because they are no longer updated. Yet they remain connected to networks,

making it easier for criminals to take advantage of unpatched

vulnerabilities.These unserviced software programs and devices are like

radioactive waste; they can’t just be left behind or thrown away. When

technology companies deprecate software, they must have a meaningful

plan to transition people away from using it. Merely saying “stop” isn’t

good enough. Companies need to ensure that people don’t lose their

investments in the software, such as source files or custom software.

The law could better facilitate such reform of the deprecation process if

it held technology companies more accountable.

Sending Sensible Signals

Design doesn’t just facilitate or hinder tasks. It also delivers information. It

communicates by sending a signal to the user of a technology.65 Through

signals, design helps define our relationships with other people and shapes

our risk calculus when interacting online. These signals affect our

expectations about how things work and the context within which we are

acting.

Data security law focuses a lot on safeguards like encryption but hardly

at all on the signals that technologies send to the people using them, like

whom a communication is coming from or how a feature works or should

be used. Yet, these signals are what encourage people to click certain links,

share personal data, or enable certain features. Designers must engineer

their systems so that the signals are understandable and that they encourage

secure rather than insecure behavior.

The wrong design can send misleading signals. People will ignore

signals if too many are sent or if not enough are sent. Poor signals can be a

major vulnerability. Fraudsters exploit confusing, inconsistent, and

ambiguous signals to trick people. In essence, poor signals make people

more gullible.

MISLEADING SIGNALS

When people use devices, there is a ton of information thrown their way

about the security risks of various activities. Unfortunately, this information

is often vague or misleading, making it difficult for people to get a clear

sense of what they should do or how vulnerable they really are.

One example is the ubiquitous padlock icon people see when they use

their browser, adjust their privacy settings on social media, and enter

authentication credentials.66 The padlock is an icon of the physical

manifestation of security—only those with the key get access to whatever it

is protecting. But what does it mean in specific contexts online? It could

mean almost anything, from the deployment of specific encryption and

authentication protocols to a general warm and fuzzy sense of “security”

similar to the comically vague statements in privacy policies that a

“company takes your privacy and security seriously.” If nothing else,

padlock icons are invitations to garner consumer trust, enticing people not

to worry about providing more personal information.67 The padlock icon

isn’t necessarily good or bad, but it is vague and companies often use it to

promise more than they deliver. Because this icon isn’t regulated, the trust

that it conveys is often false and prone to abuse.

Policymakers can play an important role in encouraging more useful

signals and discouraging vague and misleading ones. At a minimum,

regulators should provide guidance and facilitate industry coherence in

security signals for users. Regulators can also do a lot more, such as

applying federal and state laws against deceptive trade practices to

companies that use false or misleading signals. In several cases, the FTC

has alleged that technology companies using phrases like “easy to secure”

and “advanced network security” were being deceptive because their

products and services were insecure.68 Icons that invoke the concept of

security are conveying similar messages even though words aren’t being

used.

The term “security” itself also functions as a signal. Many promises

made by companies about security are as vague and empty as the padlock

icon. In other areas, the FTC has created rules to limit when companies can

use certain terms that might mislead consumers. For example, the FTC has

limited the extent to which companies can use the word “free” to describe

certain products and services.69 The use of the word “security” and icons to

represent security should be similarly scrutinized.

POORLY TIMED, TOO MANY, OR NOT ENOUGH SIGNALS

Signals not only fail when they are inconsistent, vague, or misleading.

Signals also fail based on the timing and frequency of their use. For

example, researchers at University College London conducted a study to

ascertain the effectiveness of security warnings. They found that too many

warnings desensitized people to the risks. The researchers concluded that

“security warnings in their current forms are largely ineffective, and will

remain so, unless the number of false positives can be reduced.”70 In

another study, a different group of researchers concluded that “the status

quo of warning messages appearing haphazardly—while people are typing,

watching a video, uploading files, etc.—results in up to 90 percent of users

disregarding them.”71 Researchers at Carnegie Mellon’s CyLab developed

Warning Design Guidelines, which recommended that warnings be clear,

concise, and accurate: “If too long, overly technical, inaccurate, or

ambiguous, a warning will simply be discarded and its purpose will be

lost.”72 These studies and others have repeatedly shown that security

warnings are like the porridge for Goldilocks—they have to be just right.

Too many warnings will work poorly, as will too few. Warnings at the

wrong place and time will work poorly.

Warnings should reflect the gravity of the risk. Using similar types of

warnings for low security risk and for high security risks fosters confusion;

people might begin to assume that all the warnings are for low risks and can

be ignored. Warnings must be implemented in ways to avoid being treated

like the boy who cried wolf. Moreover, the CyLab Guidelines suggest that

warnings “follow a common visual layout” because it “can be recognized

faster.”

Instead of better calibrating warnings to the risk, the opposite trend is

occurring. Some companies are now slapping warnings on all external

emails, with warnings like: EXTERNAL EMAIL—SPAM RISK. As these

warnings adorn so many harmless messages, people will become

desensitized to them.

Another problem is with spam filters. When going through the junk mail

folder to look for any legitimate emails that have been snagged, there is

often nothing to help distinguish the danger of the emails in the folder. All