Testimony of
Jerry Berman
Executive Director
Center for Democracy and Technology

Before the

House Judiciary Committee
Subcommittee on Courts and Intellectual Property


May 27, 1999



I. Overview

The Center for Democracy and Technology (CDT) is pleased to have this opportunity to testify on the issue of privacy in the online environment. CDT is a non-profit, public interest organization dedicated to developing and implementing public policies to protect and advance civil liberties and democratic values on the Internet. One of our core goals is to enhance privacy protections for individuals in the development and use of new communications technologies.

CDT focuses much of its work on the Internet because we believe that it, more than any other medium, has characteristics — architectural, economic, and social — that are uniquely supportive of democratic values. Because of its decentralized, open, and interactive nature, the Internet is the first electronic medium to allow every user to "publish" and engage in commerce. Users can reach and create communities of interest despite geographic, social, and political barriers. As the World Wide Web grows to fully support voice, data, and video, it will become in many respects a virtual "face-to-face" social and political milieu.

But while the First Amendment potential of the Internet is clear, and recognized by the Supreme Court, the impact of the Internet on individual privacy is less certain. Will the online environment erode individual privacy — building in national identifiers, tracking devices, and limits on autonomy? Or will it breathe new life into privacy — providing protections for individuals ' long held expectations of privacy?

As we move swiftly toward a world of electronic democracy, electronic commerce and indeed electronic living, it is critical to construct a framework of privacy protection that fits with the unique opportunities and risks posed by the Internet. But as Congress has discovered in its attempts to regulate speech, this medium deserves its own analysis. Laws developed to protect interests in other media should not be blindly imported. To create rules that map onto the Internet, we must fully understand the characteristics of the Internet and their implications for privacy protection. We must also have a shared understanding of what we mean by privacy. Finally we must assess how to best use the various tools we have for implementing policy — law, computer code, industry practices, and public education — to achieve the protections we seek.

II. What Makes the Internet Different?

As Congress considers crafting rules to protect privacy on the Internet, it must first understand the specific challenges to privacy posed by the Internet's functions and use.

A. Increased Data Creation and Collection

The Internet accelerates the trend toward increased information collection that is already evident in our offline world. The trail of transactional data left behind as individuals use the Internet is a rich source of information about their habits of association, speech, and commerce. When aggregated, these digital fingerprints reveal a great deal about an individual 's life. This increasingly detailed information is bought and sold as a commodity by a growing assortment of players and often sought by government.

B. The Globalization of Information and Communications

On the Internet, information and communications flow unimpeded across national borders. The Internet places the corner store, and a store three continents away, equally at the individual 's fingertips. Just as the flow of personal information across national borders poses a risk to individual privacy, citizens ' ability to transact with entities in other countries places individual privacy at risk in countries that lack privacy protections. Whether protecting citizens from fraud, limiting the availability of inappropriate content, or protecting privacy, governments are finding their traditional ability to make and effectively enforce policies challenged by the global communications medium.

C. Lack of Centralized Control Mechanisms

The Internet 's distributed architecture presents challenges for the implementation of policies. The Internet was designed without gatekeepers —there is no single entity that controls the flow of information. And as individuals and governments continually discover, the Internet offers users an unequalled ability to route around unwanted attempts to control activities and communications.

III. What Do We Mean By Privacy? And How Is It Being Eroded?

There are several core "privacy expectations" that individuals have long held vis-à-vis both the government and the private sector, the protection of which should carry over to interactions on the Internet.

A. The Expectation of Autonomy

Imagine walking through a mall where every store, unbeknownst to you, placed a sign on your back. The signs tell every other store you visit exactly where you have been, what you looked at, and what you purchased. Something very close to this is possible on the Internet.

When individuals surf the World Wide Web, they have a general expectation of anonymity, more so than in the physical world where an individual may be observed by others. Online, individuals believe that if they have not affirmatively disclosed information about themselves, then no one knows who they are or what they are doing. But, contrary to this belief, the Internet generates an elaborate trail of data detailing every stop a person makes. The individual 's employer may capture this data trail if she logged on at work, and it is captured by the Web sites the individual visits. This transactional or click stream data can provide a "profile" of an individual 's online life.

Two recent examples highlight the manner in which individuals ' expectation of autonomy is challenged. (1) The introduction of the Pentium III processor equipped with a unique identifier (Processor Serial Number) threatens to greatly expand the ability of Web sites to surreptitiously track and monitor online behavior. The PSN could become something akin to the Social Security Number of the online world — a number tied inextricably to the individual and used to validate one 's identity throughout a range of interactions with the government and the private sector. (2) The Child Online Protection Act (COPA), passed in October, requires Web sites to prohibit minors ' access to material considered "harmful to minors." Today when an individual walks into a convenience store to purchase an adult magazine they may be asked to show some identification to prove their age. Under the COPA, an individual will be asked not only to show their identification, but also to leave a record of it and their purchase with the online store. Such systems will create records of individuals ' First Amendment activities, thereby conditioning adult access to constitutionally protected speech on a disclosure of identity. This poses a Faustian choice to individuals seeking access to information — protect privacy and lose access or exercise First Amendment freedoms and forego privacy.

B. The Expectation of Fairness and Control Over Personal Information

When individuals provide information to a doctor, a merchant, or a bank, they expect that those professionals/companies will collect only information necessary to perform the service and use it only for that purpose. The doctor will use it to tend to their health, the merchant will use it to process the bill and ship the product, and the bank will use it to manage their account — end of story. Unfortunately, current practices, both offline and online, foil this expectation of privacy. Whether it is medical information, or a record of a book purchased at the bookstore, or information left behind during a Web site visit, information is routinely collected without the individual 's knowledge and used for a variety of other purposes without the individual 's knowledge — let alone consent.

The Federal Trade Commission report from last June, "Privacy Online: A Report to Congress," found that despite increased pressure, businesses operating online continue to collect personal information on the World Wide Web without providing even a minimum of consumer protection. The report looked only at whether Web sites provided users with notice about how their data was to be used; there was no discussion of whether the stated privacy policies provided adequate protection. The survey found that while 92% of the sites surveyed were collecting personally identifiable information only 14% had some kind of disclosure of what they were doing with personal data.

On May 13, 1999, the Georgetown Internet Privacy Policy Survey was released. The Survey was designed to provide an update on the state of privacy policies on the World Wide Web. The study shows that definite progress has been made in making many more Web sites privacy sensitive, but substantive privacy protections are still far from ubiquitous on the World Wide Web.

Perhaps spurred by survey upon survey documenting consumer concern and anxiety an increased number of Web sites are providing consumers with some information about what personal information is collected (44%), and how that information will be used (52%). Companies that are posting fuller information about their data handling [ 1 ] are more likely to make them accessible to consumers. Many have a link to such statements from the home page (79.7%). [ 2 ] However, on important issues such as access to personal information and the ability to correct inaccurate information, only 22% and 18% respectively of these highly trafficked Web sites provide consumers with notice. On the important issue of providing individuals with the capacity to control the use and disclosure of personal information, the survey finds that 39.5% of these busy Web sites say that consumers can make some decision about whether they are re-contacted for marketing purposes — most likely an "opt-out" — and fewer still, 25%, say they provide consumers with some control over the disclosure of data to third parties. [ 3 ]

Unfortunately, robust privacy notices remain the exception not the norm. The Survey reveals that at over 90% of the most frequently trafficked Web sites [ 4 ] consumers are not being adequately informed about how their personal information is handled. [ 5 ] At the same time the survey found that over 90% of these same busy consumer-oriented Web sites are collecting personal information. [ 6 ] In fact, the survey revealed an increase in the number of Web sites collecting sensitive information such as credit card numbers (up 20%), names (up 13.3%), and even Social Security Numbers (up 1.7%).

Thus, while many companies appear to be making an effort to address some privacy concerns, the results from the consumer perspective appear to be a quilt of complex and inconsistent statements. While progress is evident in some areas the number of sites that provide consumers with the types of notices required by the Online Privacy Alliance, the Better Business Bureau and TrustE and called for by the Federal Trade Commission and the Administration is still relatively small (9.5%).

The posting of privacy notices is not just a private sector issue. In a CDT study of federal agency Web sites, last month, we found that just over one-third of federal agencies had a "privacy notice" link from the agency 's home page. Eight other sites had privacy policies that could be found after following a link or two and on 22 of the sites surveyed we could not find a privacy policy at all.

C. The Expectation of Confidentiality

When individuals send email they expect that only the intended recipient will read it. In passing the Electronic Communications Privacy Act in 1986, Congress reaffirmed this expectation. Unfortunately, it is once again in danger.

While United States law provides email the same legal protection as a first class letter, the technology leaves unencrypted email as vulnerable as a postcard. Compared to a letter, an email message is handled by many independent entities and travels in a relatively unpredictable and unregulated environment. To further complicate matters, the email message may be routed, depending upon traffic patterns, overseas and back, even if it is a purely domestic communication. While the message may effortlessly flow from nation to nation, the privacy protections are likely to stop at the border.

Email is just one example. Today our diaries, medical records, and confidential documents are more likely to be out in the network than stored in our homes. As our wallets become "e-wallets" housed somewhere out on the Internet rather than in our back-pockets, the confidentiality of our personal information is at risk. The advent of online datebooks, and products such as Novell 's "Digital Me", and sites such as Wellmed.com [ 7 ] which invite individuals to take advantage of the convenience of the Internet to manage their lives, financial information, and even medical records raise increasingly complex privacy questions. While the real "me" has Fourth and Fifth Amendment protections from the government, the "Digital Me" is increasingly naked in cyberspace.

IV. Where Do we Go From Here?

It is clear that our policy framework did not envision the Internet as we know it today, nor did it foresee the pervasive role information technology would play in our daily lives. Our legal framework for protecting individual privacy in electronic communications, while built upon constitutional principles buttressed by statutory protections, reflects the technical and social "givens" of specific moments in history. Crafting privacy protections in the electronic realm has always been a complex endeavor. Reestablishing protections for individuals ' privacy in this new environment requires us to focus on both the technical aspects of the Internet and on the practices and policies of those who operate in the online environment.

A. The Importance of Architecture

Understanding the context is central to all effective efforts to protect privacy. While the global, distributed network environment of the Internet raises challenges to our traditional methods of implementing policies, the specifications, standards, and technical protocols that support the operation of the Internet offer a new way to implement policy decisions. By building privacy into the architecture of the Internet, we have the opportunity to advance public policies in a manner that scales with the global and decentralized character of the network. As Larry Lessig repeatedly reminds us, "(computer) code is law."

Accordingly, we must promote specifications, standards and products that protect privacy. A privacy-enhancing architecture must incorporate, in its design and function, individuals ' expectations of privacy. For example, a privacy-protective architecture would provide individuals the ability to "walk" through the digital world, browse, and even purchase without disclosing information about their identity, thereby preserving their autonomy and ensuring the expectations of privacy. A privacy-protective architecture would enable individuals to control when, how, and to whom personal information is revealed. It would also provide individuals with the ability to exercise control over how information once disclosed is, if at all, subsequently used. Finally, a privacy-protective Internet architecture would provide individuals with assurance that communications and data will be technically protected from prying eyes.

While there is much work to be done in designing a privacy-enhancing architecture, some substantial steps toward privacy protection have occurred. Positive steps to leverage the power of technology to protect privacy can be witnessed in tools like the Anonymizer, Crowds, and Onion Routing, which shield individuals ' identity during online interactions, and encryption tools such as Pretty Good Privacy that allow individuals to protect their private communications during transit. The World Wide Web Consortium 's Platform for Privacy Preferences ("P3P") is also a promising development. The P3P specification will allow individuals to query Web sites for their policies on handling personal information and to allow Web sites to easily respond. While P3P does not drive the specific practices, it is a standard designed to promote openness about information practices, to encourage Web sites to post privacy policies and to provide individuals with a simple, automated method to make informed decisions. Through settings on their Web browsers, or through other software programs, users will be able to exercise greater control over the use of their personal information.

Technologies must be a central part of our privacy protection framework, for they can provide protection across the global and decentralized Internet where law or self-regulation alone may prove insufficient.

B. Protecting the Privacy of Communications and Information

Increasingly, our most important records are not "papers" in our "houses" but "bytes" stored electronically at distant "virtual" locations for indefinite periods of time and held by third parties. The Internet, and digital technology generally, accelerate the collection of information about individuals ' actions and communications. Our communications, rather than disappearing, are captured and stored on servers controlled by third parties. Daily interactions such as our choice of articles at a news Web site, our search and purchase of an airline ticket, and our use of an online date book, such as Yahoo 's calendar, leave detailed information in the hands of third-parties. With the rise of networking and the reduction of physical boundaries for privacy, we must ensure that privacy protections apply regardless of where information is stored.

Under our existing law, there are now essentially four legal regimes for access to electronic data: 1) the traditional Fourth Amendment standard for records stored on an individual 's hard drive or floppy disks; 2) the Title III-Electronic Communications Privacy Act standard for records in transmission; 3) the standard for business records held by third parties, available on a mere subpoena to the third party with no notice to the individual subject of the record; and 4) a statutory standard allowing subpoena access and delayed notice for records stored on a remote server, such as the diary of a student stored on a university server, or personal correspondence stored on a corporate server.

As the third and fourth categories of records expand because the wealth of transactional data collected in the private sector grows and people find it more convenient to store records remotely, the legal ambiguity and lack of strong protection grows more significant and poses grave threats to privacy in the digital environment.

Congress took the first small step towards recognizing the changing nature of transactional data with amendments to the Electronic Communications Privacy Act enacted as part of the Communications Assistance for Law Enforcement Act of 1994 ("CALEA"). But the ongoing and accelerating increase in transactional data and the detail it reveals about individuals ' lives suggests that these changes are insufficient to protect privacy.

Moreover, the Electronic Communications Privacy Act must be updated to provide a consistent level of protection to communications and information regardless of where they are stored and how long they have been kept. Technologies that invite us to live online will quickly create a pool of personal data with the capacity to reveal an individual 's travels, thoughts, purchases, associations, and communications. We must raise the legal protections afforded to this growing body of detailed data regardless of where it resides on the network.

C. Establish Rules That Give Individuals Control Over Personal Information During Commercial Interactions

We must adopt enforceable standards, both self-regulatory and regulatory, to ensure that information provided for one purpose is not used or redisclosed for other purposes without the individual 's consent. All such efforts should focus on the Code of Fair Information Practices developed by the Department of Health, Education and Welfare in 1973. The challenge of implementing privacy practices on the Internet is ensuring that they build upon the medium 's real-time and interactive nature to foster privacy and that they do not unintentionally impede other beneficial aspects of the medium. Implementing privacy protections on the global and decentralized Internet is a complex task that will require new thinking and innovative approaches.

That a combination of means—self-regulation, technology, and legislation—are required is established by the most recent Georgetown Internet Privacy Policy Survey. Clearly Congress has a role to play in establishing a privacy framework for the Internet. The study, as discussed above, shows that some progress has been made in making many more Web sites privacy sensitive, but substantive privacy protections are still far from ubiquitous on the World Wide Web. Because many Web sites need baseline policy guidance and self-enforcement mechanisms while emerging may not always provide a viable remedy, we believe that legislation is both inevitable and necessary to ensure consumers' privacy on the Internet.

To achieve real privacy on the Internet, we will need more than better numbers, redoubled efforts by industry, or a legislative mantra. We will need a good-faith concerted effort by industry, consumer and privacy advocates, and policymakers to develop real and substantive answers to a number of difficult policy issues involving the scope of identifiable information, the workings of consent and access mechanisms, and the structure of effective remedies that protect privacy without adversely affecting the openness and vitality of the Internet.

As the Federal Trade Commission's rulemaking under the Children's Online Privacy Protection Act and industry's various efforts at self-regulation attest to, these issues are not easy. But armed with the findings of the Georgetown Internet Privacy Policy Survey, we believe interested parties are in a position to move forward on a three pronged approach — expanded self-regulation, work to develop and deploy privacy-enhancing technologies such as P3P, and legislation — all require a serious dialogue on policy and practice options for resolving difficult issues in this promising medium.

In its Testimony last July, the Federal Trade Commission stated that, "…unless industry can demonstrate that it has developed and implemented broad-based and effective self-regulatory programs by the end of this year, additional governmental authority in this area would be appropriate and necessary." [ 8 ] Despite the considerable effort of Congress, the Federal Trade Commission, the Administration and industry to encourage and facilitate an effective self-regulatory system to protect consumer privacy, based on the survey results we do not believe that one has yet emerged.

Last year, the Federal Trade Commission offered a legislative outline that embodied a framework, similar to the one we suggest, building upon the strengths of both the self-regulatory and regulatory processes. This year several bills have been introduced on a wide range of privacy issues. Senators Burns and Wyden, [ 9 ] and Leahy [ 10 ] have introduced proposals as have Representatives Goodlatte and Boucher, [ 11 ] and Vento. [ 12 ] We anticipate additional proposals from Senators Kohl, Torricelli, Dewine, and Hatch, and Representative Markey. Historically, for privacy legislation to be successful, it must garner the support of at least a section of the industry. To do so, it generally must build upon the work of some industry members — typically binding bad actors to the rules being followed by industry leaders — or be critically tied to the viability of a business service or product as with the Video Privacy Protection Act and the Electronic Communications Privacy Act. Several companies have staked out leadership positions on the issue of online privacy and several self-regulatory programs have formed to drive industry best practices online. In addition numerous surveys have documented that consumers are concerned about their privacy in ecommerce.

As Congress moves forward this year, we look forward to working with you and all interested parties to ensure that fair information practices are incorporated into business practices on the World Wide Web. Both legislation and self-regulation are only as good as the substantive policies they embody. As we said at the start, meaningful privacy protections that map onto the Internet requires us to resolve several critical issues. While consensus exists around at least four general principles (a subset of the Code of Fair Information Practices) — notice of data practices; individual control over the secondary use of data; access to personal information; and, security for data — the specifics of their implementation and the remedies for their violation are just beginning to be explored. We must wrestle with difficult questions: When is information identifiable? How is it accessed? How do we create meaningful and proportionate remedies that address the disclosure of sensitive medical information as well as the disclosure of inaccurate marketing data? For the policy process to successfully move forward these hard issues must be more fully resolved. The leadership of Internet-savvy members of this Committee and others will be critical as we seek to provide workable and effective privacy protections for the Internet.

D. A Privacy Protection Entity to Provide Expertise and Institutional Memory, a Forum for Privacy Research, and a Source of Policy Recommendations on Privacy Issues

The work outlined above, and the state of privacy today, all weighs in favor of creating a privacy entity within the federal government. The existing approach has hindered the development of sound policy and failed to keep pace with changes in technology. While we are pleased with the Administration 's recent appointment of Peter Swire to the Office of Information and Regulatory Affairs as the federal "privacy czar," we believe that OIRA is incapable, due to institutional constraints and a lack of autonomy, of addressing several key privacy issues. The United States needs an independent voice empowered with the scope, expertise, and authority to guide public policy. Such an entity has important roles to play on both domestic and international fronts. It would serve as the forum for collaboration with other governments, the public interest community, and the business community.

V. Conclusion

No doubt, privacy on the Internet is in a fragile state. However, there is new hope for its resuscitation. There is a special need now for dialogue. Providing a web of privacy protection to data and communications as they flow along networks requires a unique combination of tools — legal, policy, technical, and self-regulatory. Cooperation among the business community and the nonprofit community is crucial. Whether it is setting limits on government access to personal information, ensuring that a new technology protects privacy, or developing legislation — none will happen without a forum for discussion, debate, and deliberation. We thank the Committee for the opportunity to share our views and look forward to working with the members and staff and other interested parties to foster privacy protections for the Digital Age.

Footnotes

1. The report calls these "privacy policies" as compared to "information practice statements." "Privacy policies" are a more comprehensive description of a site's practices that are located in a single place and accessible through an icon or hyperlink. A site may have a "privacy policy" by this definition but still not have a privacy policy that meets the elements set out by the FTC or various industry self-regulatory initiatives for an adequate privacy policy.

2. In response to the question, "Is a Privacy Policy Notice easy to find?" surfers in the 1998 survey answered yes for approximately 1.2% of Web sites. FTC Report, Appendix C Q19.

3. This number is generated using the data from Q32 (number of sites that say they give consumers choice about having collected information disclosed to outside third parties) — 64 — and dividing it by 256 (the total survey sample (364) minus the number of sites that affirmatively state they do not disclose data to third-parties (Q29A) (69) and the number of sites that affirmatively state that data is only disclosed in the aggregate (Q30) (39)).

4. Only 9.5% of the most frequently visited Web sites and 14.7% of those that collect information had privacy policies containing critical information called for by the FTC, the Administration, and required by the Online Privacy Alliance, TrutstE and the BBB Online, about notice; choice; access; security; and contact information.

5. Last years survey found approximately 2% or Web sites that collected data, and less than 1% of all Web sites, had adequate notices.

6. 92.9% are collecting some type of personal information.

7. WellMed.com is a proprietary Online Health Management System which works by collecting personal health information from individuals, analyzing that information to develop unique health profiles which are used for a variety of purposes. One service is HealthNow! — "an online personal health record enabling secure, confidential, and private storage, management, and maintenance of health information by individuals and their families. HealthNow affords easy access of medical records from one central location anytime and anywhere the need arises."

8. Last years survey found approximately 2% or Web sites that collected data, and less than 1% of all Web sites, had adequate notices. Privacy Online: A Report to Congress, Federal Trade Commission, June 1998.

9. The Online Privacy Protection Act of 1999 (S. 809), introduced on April 15, 1999, by Senators Burns (R-MT) and Wyden (D-OR).

10. Electronic Rights for the Twenty-First Century Act of 1999 (E-RIGHTS) (S. 854), introduced on April 21, 1999 by Senator Leahy (D-VT).

11. Internet Growth and Development Act of 1999 (H.R. 1685), introduced on May 5, 1999 by Representatives Boucher (D-VA) and Goodlatte (R-VA).

12. Consumer Internet Privacy Protection Act of 1999 (H.R. 313), introduced on January 6, 1999, by Representative Vento (DFL-MN).