wildercb/PrivacyTextBooks1.2 · Datasets at Hugging Face

text stringlengths 0 118
organizations request Social Security Numbers or login information via
email, then they confuse people. Standardizing people’s expectations about
what is legitimate and what is not will help close off many ways that
hackers use to trick people.
Encouraging Balanced Security Measures
Counterintuitively, if security measures are too protective, they might lead
to bad security outcomes because people create workarounds and other
dangerous kludges that bypass protective systems entirely or weaken them.
For example, requiring two factors for authentication is generally a good
security measure. However, requiring two factors every time someone tries
to login from a particular device can be counterproductive because people
might grow frustrated and disable the service. Allowing people to check a
box to trust certain devices after using two-factor authentication for the first
time will help ease the inconvenience. In another example, which we
discussed earlier, if people are forced to select passwords that are too long
and complex, they will write them down on sticky notes near their
computers, which opens up another major security vulnerability.63
Good public health frameworks are careful not to ask too much of
people and instead seek to find the right balance between protection and
feasibility to ensure measures can be implemented across an entire
population.64 It seems commonsensical, yet organizations regularly design
their systems in ways that make us and our data less secure.
ELIMINATING IMPOSSIBLE SECURITY PRACTICES AND ADVICE
Several common security measures are not practical for people to do. The
law should discourage the use of security measures that fail in this way.
Instead, however, the law perversely encourages such measures. Regulators
often look superficially at an organization’s security, faulting organizations
that don’t use certain measures. Thus, security professionals might follow
the same old faulty security measures because they fear that regulators will
zing them for missing measures on the checklist. Regulators and the law
must ensure that they stop playing a role in furthering this pathology.
If these measures were to be abandoned, what would work better? An
improved strategy to help people with passwords is to use password
managers to store them electronically in one account. Password manager
services help people create strong, unique passwords, warn them when their
passwords are bad, and save them the trouble of remembering the
potentially thousands of passwords they use to access their accounts. Even
these services are typically secured, at least partially, by a password
themselves. Although these services are typically a better way of generating
and keeping track of your passwords, like anything on the Internet, they are
not without risk.
Without a password manager, we recommend different passwords at
work and for personal accounts. But beyond this, we don’t recommend a
different password for every account. People can have hundreds of
accounts, and it’s not possible to remember hundreds of passwords. It is
easy to find cases of password re-use and chastise people, but that’s asking
the impossible and scapegoating. What would be far better is to ensure that
different passwords are used when it really counts—such as work accounts
—and not as a matter of course.
Regarding not writing passwords down, it is impossible to remember
many passwords. People must write them down. Instead of exhorting
people never to write passwords down, the advice should be about where to
write and keep them. People should not put them near the workplace
computer or carry them in wallets. But suppose a person kept a list of
passwords in a dresser drawer. Conceivably, a hacker could break into the
person’s home, rummage everywhere, and find the passwords in the dresser.
But this is unlikely and is low risk. Of course, it is still a risk, and rigid
approaches to security would try to stamp out this practice to eliminate the
risk. The problem is that trying to reduce this low risk might ultimately
induce people to engage in more risky behavior.
Good security is like good parenting. Parents soon learn to pick their
battles. They can’t win them all. Designing security practices to reduce all
risks, even the small ones and trivial ones, can weaken human responses to
risks that are more severe. The perfect becomes the enemy of the good.
TWO-FACTOR AUTHENTICATION
In security, there are other solutions to authentication problems and
methods of authentication that can be used if organizations moved away
from passwords alone. Many relatively cheap and easy-to-deploy methods
can be used to protect against different kinds of attacks on credentials.
One means of authentication that accommodates human limitations is
two-factor authentication. The essence of two-factor authentication is
simple. To login, you must have something you know (usually a password),
as well as one additional factor, usually something you have (e.g., your
cellphone).
Two-factor authentication is promising because it has already been
deployed by major companies, it protects against many different kinds of
offline attacks, and it can leverage a technology that most people already
constantly carry around—their cellphone. Two-factor authentication is a
good way to protect against both online and offline attacks. While two-
factor authentication remains vulnerable to specialized phishing and
malware-based attacks, those vulnerabilities are relatively narrow and
typically require the fraudster to already have a person’s username and
password.
Two-factor authentication is not a silver bullet that addresses all the
problems with passwords. There are no useful silver bullets in data security.
But there are some measures that will reduce risk.
Although many of these techniques are widely available and
inexpensive, they are often not used. Change is not likely to happen fast
enough without regulatory intervention. Perhaps a nudge, maybe a gentle
push, or maybe a forceful shove will be needed. The law should promote
security measures like two-factor authentication that are practical in light of
human behavior.
ANTICIPATING HUMAN PROBLEMS
Companies often deprecate their software and Internet-connected devices
and stop supporting them. They often ignore how attached people are to the
tools they use. They tell people to simply stop using them. This isn’t
realistic. People won’t just stop using old devices and programs; they often
have too much invested in the software to just give it up at the command of
a tech company. The difficulty in moving on is a data security problem.
Time and again, hackers have targeted deprecated software and devices,

organizations request Social Security Numbers or login information via

email, then they confuse people. Standardizing people’s expectations about

what is legitimate and what is not will help close off many ways that

hackers use to trick people.

Encouraging Balanced Security Measures

Counterintuitively, if security measures are too protective, they might lead

to bad security outcomes because people create workarounds and other

dangerous kludges that bypass protective systems entirely or weaken them.

For example, requiring two factors for authentication is generally a good

security measure. However, requiring two factors every time someone tries

to login from a particular device can be counterproductive because people

might grow frustrated and disable the service. Allowing people to check a

box to trust certain devices after using two-factor authentication for the first

time will help ease the inconvenience. In another example, which we

discussed earlier, if people are forced to select passwords that are too long

and complex, they will write them down on sticky notes near their

computers, which opens up another major security vulnerability.63

Good public health frameworks are careful not to ask too much of

people and instead seek to find the right balance between protection and

feasibility to ensure measures can be implemented across an entire

population.64 It seems commonsensical, yet organizations regularly design

their systems in ways that make us and our data less secure.

ELIMINATING IMPOSSIBLE SECURITY PRACTICES AND ADVICE

Several common security measures are not practical for people to do. The

law should discourage the use of security measures that fail in this way.

Instead, however, the law perversely encourages such measures. Regulators

often look superficially at an organization’s security, faulting organizations

that don’t use certain measures. Thus, security professionals might follow

the same old faulty security measures because they fear that regulators will

zing them for missing measures on the checklist. Regulators and the law

must ensure that they stop playing a role in furthering this pathology.

If these measures were to be abandoned, what would work better? An

improved strategy to help people with passwords is to use password

managers to store them electronically in one account. Password manager

services help people create strong, unique passwords, warn them when their

passwords are bad, and save them the trouble of remembering the

potentially thousands of passwords they use to access their accounts. Even

these services are typically secured, at least partially, by a password

themselves. Although these services are typically a better way of generating

and keeping track of your passwords, like anything on the Internet, they are

not without risk.

Without a password manager, we recommend different passwords at

work and for personal accounts. But beyond this, we don’t recommend a

different password for every account. People can have hundreds of

accounts, and it’s not possible to remember hundreds of passwords. It is

easy to find cases of password re-use and chastise people, but that’s asking

the impossible and scapegoating. What would be far better is to ensure that

different passwords are used when it really counts—such as work accounts

—and not as a matter of course.

Regarding not writing passwords down, it is impossible to remember

many passwords. People must write them down. Instead of exhorting

people never to write passwords down, the advice should be about where to

write and keep them. People should not put them near the workplace

computer or carry them in wallets. But suppose a person kept a list of

passwords in a dresser drawer. Conceivably, a hacker could break into the

person’s home, rummage everywhere, and find the passwords in the dresser.

But this is unlikely and is low risk. Of course, it is still a risk, and rigid

approaches to security would try to stamp out this practice to eliminate the

risk. The problem is that trying to reduce this low risk might ultimately

induce people to engage in more risky behavior.

Good security is like good parenting. Parents soon learn to pick their

battles. They can’t win them all. Designing security practices to reduce all

risks, even the small ones and trivial ones, can weaken human responses to

risks that are more severe. The perfect becomes the enemy of the good.

TWO-FACTOR AUTHENTICATION

In security, there are other solutions to authentication problems and

methods of authentication that can be used if organizations moved away

from passwords alone. Many relatively cheap and easy-to-deploy methods

can be used to protect against different kinds of attacks on credentials.

One means of authentication that accommodates human limitations is

two-factor authentication. The essence of two-factor authentication is

simple. To login, you must have something you know (usually a password),

as well as one additional factor, usually something you have (e.g., your

cellphone).

Two-factor authentication is promising because it has already been

deployed by major companies, it protects against many different kinds of

offline attacks, and it can leverage a technology that most people already

constantly carry around—their cellphone. Two-factor authentication is a

good way to protect against both online and offline attacks. While two-

factor authentication remains vulnerable to specialized phishing and

malware-based attacks, those vulnerabilities are relatively narrow and

typically require the fraudster to already have a person’s username and

password.

Two-factor authentication is not a silver bullet that addresses all the

problems with passwords. There are no useful silver bullets in data security.

But there are some measures that will reduce risk.

Although many of these techniques are widely available and

inexpensive, they are often not used. Change is not likely to happen fast

enough without regulatory intervention. Perhaps a nudge, maybe a gentle

push, or maybe a forceful shove will be needed. The law should promote

security measures like two-factor authentication that are practical in light of

human behavior.

ANTICIPATING HUMAN PROBLEMS

Companies often deprecate their software and Internet-connected devices

and stop supporting them. They often ignore how attached people are to the

tools they use. They tell people to simply stop using them. This isn’t

realistic. People won’t just stop using old devices and programs; they often

have too much invested in the software to just give it up at the command of

a tech company. The difficulty in moving on is a data security problem.

Time and again, hackers have targeted deprecated software and devices,