wildercb/PrivacyTextBooks1.2 · Datasets at Hugging Face

text stringlengths 0 118
them to do anything with data on their own. But that can come at a cost
because these control measures can be oppressive and counterproductive,
changing the culture of an organization and making it too rigid, less free,
and less trusting.
Good data security involves determining the appropriate level of risk.
How much risk is appropriate? There is no one right answer to this
question. Thoughtful data security decisions involve judgments about
balancing many considerations. Data security is much more complicated,
nuanced, and creative than merely checking items off a list. Although
lawmakers often treat data security as a science with objective correct
answers, it is much more of an art.9
Recall that the main goals of cybersecurity are most commonly
conceptualized as protecting the “confidentiality, integrity, and availability”
of data—the CIA triad.10 While these values are useful as specific goals for
policies and practices, they don’t capture all of the dimensions of managing
the risks involved with personal data. There is much more to data protection
than maintaining confidentiality, such as addressing how personal data is
shared, used, and stored.
If data security is thought of primarily in terms of risk management, it
means the goals of confidentiality, integrity, and availability must be
consistently balanced against the goals of information systems and the
societal goals like market participation, self-expression, socialization,
public health, and even democratic self-governance.
Here is where data security law can go astray. At the organizational
level, over-enforcement against breached companies can skew risk
management. At the systems level, notification requirements and safeguard
rules don’t address the need to allocate and calibrate responsibility for
protection across groups of actors who contribute to insecurity. A different
approach is needed.
A NEW ROLE FOR DATA SECURITY LAW
Although the current law is not adequately addressing the challenge of data
security, there is an important role for the law to play. But lawmakers must
take a fundamentally different approach. Currently, the law operates
primarily by worsening the consequences of a breach. The hope is that
intensifying the pain of breaches will deter breaches by making
organizations take more steps to avoid them. But extreme deterrence is the
wrong goal. As we discussed above, the goal is to encourage optimal risk
management, not to avoid breaches at all costs. Because organizations that
suffer breaches are not the only cause of the breaches or the only actors that
can affect the risk, the law should facilitate a better management of the risk
across the entire data ecosystem.
Data security law should take on the role of promoting structural change
in the data ecosystem. The law should establish who is responsible for
what. It should set up the right incentives; it should help shape the system in
which everything operates. The key to the law’s success is to look beyond
the breach to the whole data ecosystem. As we will discuss later, data
breaches are the product of many actors in the data ecosystem, and the law
currently only focuses on a fraction of them.
An Analogy to Public Health
Data security could learn a great deal by looking through the lens of public
health. The two fields actually use similar terms and concepts. Data security
professionals speak to the “health” of their network.11 They refer to
mapping and monitoring their data flows as good “data hygiene.”12
“Viruses” threaten both public health and the security of networks and data.
The language of health in these settings is both appropriate and
underappreciated by law and policy makers. The Oxford English Dictionary
defines health as “soundness of body; that condition in which its functions
are duly and efficiently discharged.”13 If we consider data networks as a
body that must function efficiently, then lawmakers might benefit from an
alternative approach to data security law and policy that looks beyond the
health of individuals (in our case, data holders) and addresses the health of
an entire populous. The field of public health looks at “the health of the
population as a whole, especially as monitored, regulated, and promoted by
the state (by provision of sanitation, vaccination, etc.).”14 As Wendy Parmet
notes: “[I]nfectious epidemics show that the health of an individual
depends, to a great degree, on the health of others. . . . An individuals’ risk
of becoming ill depended in uncertain ways on the steps that the community
took and the environment and conditions in which the individual lived.”15
The key lesson from public health is interdependency—we have to focus on
the whole system, not play whack-a-mole with specific individuals.
Similar to combating disease epidemics, addressing the data security
epidemic requires collective action. The security of our data largely
depends on factors that are outside of our control and even significantly
outside the control of the organizations that are responsible for storing it
securely. Our approach here is inspired by Deirdre Mulligan and Fred
Schneider’s conceptualization of cybersecurity as a public good.16 The
scholars noted that “Public health and cybersecurity both aim to achieve a
positive state (health or security) in a loosely affiliated but highly
interdependent network.”17 Adam Shostack has also articulated a vision of
“cyber public health” that would support public health equivalents of
institutions and practices for cybersecurity.18 “Security experts rarely give
advice on the level of ‘wash your hands’,” Shostack notes. “Their advice is
rarely consistent with other experts, or the public. People are naturally
confused and give up.” He further argues that unlike in public health, where
public health institutions provide useful statistics and guidance, there are
“few equivalents in the world of cybersecurity.”19
Another important feature of looking at data security from a health and
wellness perspective is that, just like illnesses are inevitable, so too are
security breaches. Derek Bambauer has argued that “focusing efforts
principally on preventing cyberattacks is misguided: perfect security is
impossible, and even attaining good security is extraordinarily difficult.
Instead, cybersecurity regulation should concentrate on mitigating the
damage that successful attacks cause.”20
The Who, What, When, and How of Data Security Law
Data security law jumps in at the wrong time, is not looking to all the
responsible actors, is not addressing the practices that most lead to poor
security, and is not working in the right way. It is failing in four key
dimensions—who, what, when, and how.
WHO
The law loves to pummel the breached organization. Of course, this is an

them to do anything with data on their own. But that can come at a cost

because these control measures can be oppressive and counterproductive,

changing the culture of an organization and making it too rigid, less free,

and less trusting.

Good data security involves determining the appropriate level of risk.

How much risk is appropriate? There is no one right answer to this

question. Thoughtful data security decisions involve judgments about

balancing many considerations. Data security is much more complicated,

nuanced, and creative than merely checking items off a list. Although

lawmakers often treat data security as a science with objective correct

answers, it is much more of an art.9

Recall that the main goals of cybersecurity are most commonly

conceptualized as protecting the “confidentiality, integrity, and availability”

of data—the CIA triad.10 While these values are useful as specific goals for

policies and practices, they don’t capture all of the dimensions of managing

the risks involved with personal data. There is much more to data protection

than maintaining confidentiality, such as addressing how personal data is

shared, used, and stored.

If data security is thought of primarily in terms of risk management, it

means the goals of confidentiality, integrity, and availability must be

consistently balanced against the goals of information systems and the

societal goals like market participation, self-expression, socialization,

public health, and even democratic self-governance.

Here is where data security law can go astray. At the organizational

level, over-enforcement against breached companies can skew risk

management. At the systems level, notification requirements and safeguard

rules don’t address the need to allocate and calibrate responsibility for

protection across groups of actors who contribute to insecurity. A different

approach is needed.

A NEW ROLE FOR DATA SECURITY LAW

Although the current law is not adequately addressing the challenge of data

security, there is an important role for the law to play. But lawmakers must

take a fundamentally different approach. Currently, the law operates

primarily by worsening the consequences of a breach. The hope is that

intensifying the pain of breaches will deter breaches by making

organizations take more steps to avoid them. But extreme deterrence is the

wrong goal. As we discussed above, the goal is to encourage optimal risk

management, not to avoid breaches at all costs. Because organizations that

suffer breaches are not the only cause of the breaches or the only actors that

can affect the risk, the law should facilitate a better management of the risk

across the entire data ecosystem.

Data security law should take on the role of promoting structural change

in the data ecosystem. The law should establish who is responsible for

what. It should set up the right incentives; it should help shape the system in

which everything operates. The key to the law’s success is to look beyond

the breach to the whole data ecosystem. As we will discuss later, data

breaches are the product of many actors in the data ecosystem, and the law

currently only focuses on a fraction of them.

An Analogy to Public Health

Data security could learn a great deal by looking through the lens of public

health. The two fields actually use similar terms and concepts. Data security

professionals speak to the “health” of their network.11 They refer to

mapping and monitoring their data flows as good “data hygiene.”12

“Viruses” threaten both public health and the security of networks and data.

The language of health in these settings is both appropriate and

underappreciated by law and policy makers. The Oxford English Dictionary

defines health as “soundness of body; that condition in which its functions

are duly and efficiently discharged.”13 If we consider data networks as a

body that must function efficiently, then lawmakers might benefit from an

alternative approach to data security law and policy that looks beyond the

health of individuals (in our case, data holders) and addresses the health of

an entire populous. The field of public health looks at “the health of the

population as a whole, especially as monitored, regulated, and promoted by

the state (by provision of sanitation, vaccination, etc.).”14 As Wendy Parmet

notes: “[I]nfectious epidemics show that the health of an individual

depends, to a great degree, on the health of others. . . . An individuals’ risk

of becoming ill depended in uncertain ways on the steps that the community

took and the environment and conditions in which the individual lived.”15

The key lesson from public health is interdependency—we have to focus on

the whole system, not play whack-a-mole with specific individuals.

Similar to combating disease epidemics, addressing the data security

epidemic requires collective action. The security of our data largely

depends on factors that are outside of our control and even significantly

outside the control of the organizations that are responsible for storing it

securely. Our approach here is inspired by Deirdre Mulligan and Fred

Schneider’s conceptualization of cybersecurity as a public good.16 The

scholars noted that “Public health and cybersecurity both aim to achieve a

positive state (health or security) in a loosely affiliated but highly

interdependent network.”17 Adam Shostack has also articulated a vision of

“cyber public health” that would support public health equivalents of

institutions and practices for cybersecurity.18 “Security experts rarely give

advice on the level of ‘wash your hands’,” Shostack notes. “Their advice is

rarely consistent with other experts, or the public. People are naturally

confused and give up.” He further argues that unlike in public health, where

public health institutions provide useful statistics and guidance, there are

“few equivalents in the world of cybersecurity.”19

Another important feature of looking at data security from a health and

wellness perspective is that, just like illnesses are inevitable, so too are

security breaches. Derek Bambauer has argued that “focusing efforts

principally on preventing cyberattacks is misguided: perfect security is

impossible, and even attaining good security is extraordinarily difficult.

Instead, cybersecurity regulation should concentrate on mitigating the

damage that successful attacks cause.”20

The Who, What, When, and How of Data Security Law

Data security law jumps in at the wrong time, is not looking to all the

responsible actors, is not addressing the practices that most lead to poor

security, and is not working in the right way. It is failing in four key

dimensions—who, what, when, and how.

WHO

The law loves to pummel the breached organization. Of course, this is an