Director, Electronic Privacy Information Center
Adjunct Professor, Georgetown University Law Center
Subcommittee on Courts and Intellectual Property
House Judiciary Committee
U.S. House of Representatives
March 26, 1998
Public opinion polls show that privacy is the number one concern of Internet users. Everyone is aware that a great deal of personal information is collected, and that virtually no meaningful protections are in place.
In the McVeigh-AOL case, a person almost lost his job because of information that was improperly disclosed by an online service provider. An amendment to Electronic Communications Privacy Act could help prevent similar incidents in the future. But the example is just one of the many privacy risks that people using the Internet today face.
The Internet lacks adequate privacy protection. A survey by the Electronic Privacy Information Center in 1997 of the 100 top web sites found that less than half had privacy policies, and those with policies offered little real protection. Still, anonymity plays a critical role in online privacy as it gives individuals the ability to control the disclosure of their identity.
Even though the Internet is a very new communications environment, the commitment to establish privacy protection by law in the United States is long-standing. The US developed important legal safeguards to protect the privacy of communications and established the fundamental approach to the protection of personal information -- generally described as "Fair Information Practices" -- that make clear the responsibilities of organizations that collect data and the rights of individuals who give up personal information.
We are at the beginning of a long and difficult period for the protection of privacy in this country. Technology is racing ahead. Our laws and institutions are lagging far behind. The level of public concern about privacy is growing. There is much work to be done.
My name is Marc Rotenberg. I am the Director of the Electronic Privacy Information Center, a non-partisan research organization in Washington, DC. I am an adjunct professor at Georgetown University Law Center and Senior Lecturer at the Washington College of Law. I am also editor, with Philip E. Agre, of Technology and Privacy: The New Landscape (MIT Press 1997).
I appreciate the opportunity to testify before the Subcommittee today. I'd like to thank the Subcommittee for holding this hearing and also for your ongoing work in support of Representative Goodlatte's SAFE bill that would help reform our nation's policy on encryption.
The growing concern about the loss of privacy on the Intenet was made clear earlier this year when the Navy began discharge proceedings against a decorated sailor based on personal information about the sailor disclosed by America Online. A Navy investigator, suspecting that Mr. McVeigh might be in violation of the "Don't Ask, Don't Tell" policy, obtained information that linked Mr. McVeigh's "screen name," which was not his actual identity, with his real identity. Once the connection was established, the discharge proceeding began.
The McVeigh-AOL case raised a complicated set of legal issues. The AOL Terms of Service agreement specifically prohibited this disclosure. But a civil action against the company would not mean reinstatement by the Navy. The disclosure also appeared to violate the Electronic Communications Privacy Act, but the statute is ambiguous about the remedies available to victims of such disclosure.
Mr. McVeigh filed suit against Navy Secretary John Dalton in federal court. Judge Stanley Sporkin found that the Navy had violated the "Don't Ask, Don't Tell policy" when it pursued the investigation. In the course of the decision, Judge Sporkin also considered whether the Navy violated the Electronic Communications Privacy Act. The opinion is a little less clear on this point. Judge Sporkin said the investigation undertaken by the Navy was "likely illegal" under the ECPA because the Navy investigator failed to obtain a warrant before he sought personal information from America Online about Mr. McVeigh. The government contended that the obligation to comply with ECPA fell not on the government actor but rather on the online service provider.
Judge Sporkin said that the statute read as a whole made clear the intent to regulate the conduct of government agents. He found that even if the relevant provision did not apply to the actions of government (18 USC § 2703), "it is elementary that information obtained improperly can be suppressed where an individual's rights have been violated." Judge Sporkin concluded "in these days of 'big brother,' where through technology and otherwise the privacy interests of individuals from all walks of life are being ignored or marginalized, it is imperative that statues explicitly protecting these rights be strictly observed."
The McVeigh case is critical for several reasons. First, it makes clear that privacy violations have real consequences. Mr. McVeigh's life was forever changed by the decision of America Online to disclose personal information about him to his employer. Second, the case shows the shortcomings of contractual solutions. Even with a very clear contract provision detailing when personal information may be disclosed, the Navy investigator was still able to obtain personal information about Mr. McVeigh. Third, the case shows that we are all becoming increasingly dependent on these new services to safeguard our privacy. America Online today has more than eleven million subscribers.
Mr. McVeigh's case, because the improper disclosure of information was so well documented, received national attention. But there are many other people in this country who face similar privacy risks, whose names will never be known. Indeed, they themselves may never know that information about them was improperly disclosed.
What is Privacy?
In some respects, the McVeigh case appears complicated. AOL didn't actually disclose personal information about Mr. McVeigh, such as an unlisted phone number or a Social Security Number. Rather, the company disclosed information that linked his actual identity to an assumed identity. The Internet raises many privacy issues that seem novel or unusual:
As complicated as these examples may seem, the basic privacy analysis is not so difficult. The premise that virtually all privacy law and policy is based on is the belief that when individuals give up personally identifiable information to organizations, the organizations take on some obligation and the individuals are granted some rights. We call these responsibilities and rights "Fair Information Practices."
The critical elements of Fair Information Practices include:
You will find this approach to privacy protection in virtually all of the privacy laws in the United States, including many of the recent statutes that address new technologies, such as the subscriber privacy provision in the Cable Act of 1984, the Electronic Communications Privacy Act of 1986, the Video Privacy Protection Act of 1998 (video tape rentals), the Telephone Consumer Protection Act of 1991 (auto-dialers and junk faxes), and even the CPNI rules contained in the Telecommunication Reform Act of 1996 (customer billing information).
To be effective, Fair Information Practices must be enforced and must provide redress. It is not enough to say what a policy is without providing a means to enforce the policy. That is why voluntary guidelines, professional standards, and codes of conduct that are based on Fair Information Practices do not necessarily provide significant privacy protection.
There are also some novel issues. One very interesting and very important policy question is brought about by the development of new technologies that make it possible to protect privacy in ways we had not previously imagined. Traditionally, we understood that technology was a threat to privacy and that it was the proper role of government to restrict the use of techniques that might intrude on privacy. But now we see in such techniques as public key encryption and anonymous payment schemes the opportunity to develop new means to limit the disclosure of personal information.
The critical question then becomes what role government should play in promoting, regulating, or restricting techniques such as encryption that allow individuals to protect personal information. In the United States this debate has largely been framed in terms of the need to balance the interests of privacy and commerce against the concerns of law enforcement and national security. But in most other parts of the world that have looked at this issue, there is a very different view. Many governments believe that these new technologies should be promoted and that efforts to impose controls for law enforcement purposes are short-sited and will ultimately prove futile.
In my view, privacy in the information age means both the extension of Fair Information Practices to new information environments and the active promotion of techniques, often based on encryption, to protect the disclosure of personal information. This is the fundamental policy goal.
Understanding the Problem of Privacy on the Internet
To understand the problem of privacy on the Internet in more detail, EPIC conducted a survey of the top 100 web sites in the summer of 1997. It was the first comprehensive survey of Internet privacy. We looked at the policies and practices actually in place on the most popular web sites. For each site, we checked whether personally identifiable information was collected, whether a notice describing privacy polices was displayed, whether the policy was adequate, and similar questions.
We found that about half of the sites that we surveyed collected personal information. This was typically done for on-line registration, surveys, user profiles and order fulfillment. Seventeen sites had privacy notices or statements, but the policies were often not easy to locate and some policies we could only find after we registered at the site.
One of the most interesting findings in our survey was that anonymity was largely respected by the websites. Most websites allow users to visit and receive information -- about products, or news, or almost anything else you can find on the Internet -- without collecting personal information.
In the conclusions of our report we said that:
We closed with the warning "surfer beware" because we concluded that there was simply too little privacy protection on the Internet for users to feel secure, and we hoped stronger privacy standards would be developed.
This month the Federal Trade Commission is conducting a similar survey of 1,200 web sites. I suspect that the FTC will find that a growing numbers of web sites do now have privacy polices. But whether those policies are meaningful or provide any redress to users of these services remains unclear. It is worth noting that America Online has one of the most comprehensive and detailed privacy policies of any company operating on the Internet today. And still Timothy McVeigh almost lost his job.
History of Communications Privacy
One of the great achievements of the American legal system has been our strong commitment to protecting the privacy of personal communications. You can trace this history back at least as far as Benjamin Franklin, who in establishing the national postal service recognized the need to enact federal law to protect the privacy of communications.
But it was not until 1928 that the Supreme Court had its first brush with the question of whether our Bill of Rights, drafted in the eighteenth century, would apply to the new communications technologies of the twentieth century and beyond. The case concerned a highly successful bootlegging operation in the Pacific Northwest operated by Ralph Olmstead. Federal agents began an extensive surveillance operation that lasted for more than five months. They had no recording devices, so they wrote down what they heard. Sometimes they relayed their recollections of conversations to a stenographer. In the end, they compiled more than 775 pages that they brought to court. The issue was whether the Fourth Amendment warrant requirement would be applied to this new investigative technique. The trial court let the evidence in, over the objection of Mr. Olmstead, and the appeals court affirmed.
When the case finally reached the Supreme Court Chief Justice William Taft wrote a detailed opinion that focused on the absence of a physical search, of the type proscribed by the Fourth Amendment, and concluded that the evidence was admissible. The Court held that the Fourth Amendment simply did not apply to this new form of communication.
But there were two important dissents. Justice Holmes called the matter a "dirty business" because the federal agents had violated a Washington state law that prohibited wiretapping to obtain the evidence. He voted to reverse.
Justice Brandeis also dissented. His opinion was not so much about the illegal acts of federal agents; he was more interested in the question of how the Fourth Amendment and our Constitution generally, should apply to these new communication technologies. He wrote, in one of the most famous phrases in American law, that the makers of our Constitution "sought to protect Americans in their beliefs, their thoughts, their emotions and their sensations. They conferred, as against the Government, the right to be let alone -- the most comprehensive of all rights and the right most valued by civilized men." Brandeis's dissent in Olmstead reminds us that the protection of privacy is at the heart of our system of ordered liberty and that that law is an evolving process.
The Supreme Court eventually adopted Justice Brandeis's view and decided in 1967 that the Fourth Amendment did indeed apply to telephone communications. Following the Katz decision and a related case, Berger v. New York, the Congress set out in 1968 to establish a framework to allow electronic wiretapping only under the most limited circumstances. The Congress made clear at that time that wiretapping was to be an investigative means of "last resort."
While some have said that Title III makes clear that the police have the right to wiretap telephone communications when a court order is obtained, I believe the better view of the Act is that it ensured that electronic surveillance would be brought within strict Fourth Amendment requirements. In other words, our federal wiretapping statute was intended to limit this investigative technique to the narrowest circumstances.
Since 1967 there have been a number of significant developments in the law of communications privacy. In 1978 the Congress passed the Foreign Intelligence Surveillance Act to deal with the difficult problem of wiretapping of foreign agents. The Supreme Court had left open the question in the Katz case of whether the Fourth Amendment should apply to national security cases. The Congress resolved this question with the FISA in 1978, establishing a Title III-like framework, albeit with more secrecy and less accountability.
In the mid-1980s the growth of the Internet and new communications services was apparent. People were using desktop computers and sending messages to one another by means of electronic mail. Questions about the appropriate standards for government searches were arising. In response, Congress amended Title III and enacted the Electronic Communications Privacy Act, which extended privacy protection to stored electronic communications.
The next significant development came in 1994 when Congress passed the Communications Assistance for Law Enforcement Act (CALEA), a measure commonly referred to as "digital telephony." CALEA gave the Department of Justice the authority to set technical standards for the nation's telephone system in an attempt to ensure the ongoing viability of wiretapping.
Many said at the time that the measure was considered that it was a mistake to pass such legislation, not only because it was a fundamental change in the law's approach to electronic surveillance and police powers generally, but also that the bill would be impractical and ultimately unworkable.
For better or worse, this predication seems now to be correct. The FBI and the telephone industry are mired in endless debates about implementing the legislation, the estimated costs are far beyond the initial authorization, the technology innovations continue, and the CALEA policy has slowed the adoption of technical methods, such as encryption, that could make our communications network more secure and reduce the risk of crime. Moreover, our government is now in the unfortunate position of urging other nations to develop more extensive surveillance capabilities.
I hope at some point in the future the Judiciary Committee will have the opportunity to revisit CALEA and to consider whether this is still a sensible policy initiative.
The Role of Government
The United States was for many years a leader in efforts to protect personal privacy. Justice Brandeis wrote a famous law review article on the right to privacy in the late nineteenth century that established the legal claim in this country and elsewhere. The privacy right came to be described as the "American tort."
Many other countries joined the US effort to firmly establish this right following the end of the Second World War. The Universal Declaration of Human Rights was adopted and the right of privacy was made explicit in the constitutions of many governments.
The United States continued to lead in the modern era of privacy protection with passage of the Fair Credit Reporting Act in 1970 and then with the Privacy Act of the 1974 that provided comprehensive privacy protection for records held by the federal government.
But our lead has slipped, and we are now viewed by many as falling behind in the effort to protect this critical right. The Administration's own record on privacy protection has been very poor. Not only has the White House resisted calls from long-time trading partners and allies to develop stronger privacy measures, it has actively opposed efforts by other governments to extend privacy rights to their own citizens. This combined with the Administration's attempt to extend techniques for electronic surveillance has placed the United States in the unfortunate position of promoting state surveillance as other governments are trying to establish privacy protection.
The sharp contrast in our government's approach to privacy issues, when compared with other governments, can be understood by considering the significance of the date "October 1998." In Europe that is the date when the European Data Directive goes into force. It is a comprehensive privacy measure that establishes rights for citizens and recognizes that privacy protection will remain critical for the information economy. It is the result of many years of hard work, negotiation, and commitment by lawmakers.
In this country, in October 1998, we will mark the date when the Communications Assistance for Law Enforcement Act is expected to be operational. That is the law, as I have noted, that requires telephone companies to try to protect electronic interception in the nation's telephone system. We are pursuing elaborate and expensive policies for national communications surveillance as other countries are struggling with the issue of how to protect the privacy rights of their citizens.
We are today not only behind the curve in developing sensible privacy polices, but we are largely out of step with the rest of the world. Lacking the formal means to develop privacy policies and to respond to public concerns, we have left the law enforcement community and the marketing industry to determine how much privacy there will be in the future. The result is not surprising -- there is growing public concern about the loss of privacy and a widening gap between the problems we face and the solutions we should pursue.
Simply stated, our policy is backward. We impose government controls on techniques to protect privacy, where market-based solutions are preferable. And we leave privacy problems to the market, where government involvement is required.
Today the calls for government action to protect privacy are unambiguous. The most recent Harris poll found that a majority of those polled found that privacy is the main reason that people are staying off of the Internet. They want legislation now to protect privacy on the Internet. According to the BusinessWeek/Harris poll, 53% believe that "Government should pass laws now for how personal information can be collected and used on the Internet." Of those polled, 23% said "government should recommend privacy standards for the Internet but not pass laws at the time." Only 19% believe that the government "should let groups develop voluntary privacy standards but not take any action now unless real problems arise."
The Harris/BusinessWeek poll is consistent with other polls that have asked similar questions about privacy and the Internet. Contrary to the popular view that Internet users oppose all form of government action, when it comes to matters of privacy, they believe new laws are necessary.
Where once individual consent was central to the disclosure of personal information, now the focus is on individual choice for a range of disclosures. Where privacy techniques focused on the means to protect identity, now the focus is on means to obtain information. Many of the techniques that are put forward as "technical solutions" --- such as the Open Profiling Standard, the P3P and Trustee --- will make it easier, not more difficult, to obtain information from individuals using the Internet. Something is clearly amiss.
It is time to reestablish support for Fair Information Practices, to make clear that organizations that collect information have responsibilities, and that individuals who give up information have rights. The principles are well established in our legal tradition. Privacy protection should not end where the Internet begins.
Amend the Electronic Communications Privacy Act
Congress should specifically consider expanding the scope of privacy provided to subscriber information under Section 2703 of ECPA. Currently, the statute only prohibits the disclosure of such data to "governmental entities" unless they obtain legal process authorizing the disclosure. This prohibition should be extended to the disclosure of subscriber information to any third party. One of the reasons why the Navy was able to obtain information concerning Mr. McVeigh from AOL is that ECPA places no restrictions on service providers unless the requester identifies himself as a government agent, which the Navy investigator failed to do. Further, the current statutory regime fails to recognize that significant harm can result from the disclosure of personal information to non-governmental actors. Had Mr. McVeigh been a private sector employee, ECPA would have provided absolutely no protection, despite the fact that he could have lost his job in much the same way. Any requester should be required to provide legal authorization before receiving personal information from a service provider.
With respect to governmental access, ECPA should be amended to prohibit the use as evidence of information obtained in violation of Section 2703, in the same way that Section 2515 prohibits the use of illegally obtained wire or oral communications.
Finally, the civil action provision contained in Section 2707 should be amended to make clear that a cause of action will lie against a governmental entity that obtains information in violation of Section 2703.
Support Passage of Internet Privacy Bill and the Children Privacy Bill
The Consumer Internet Privacy Protection Act of 1997 (HR 98) would prevent an "interactive computer service" from disclosing to a third party a subscriber's personal information without that individual's written content. This is a good starting point but will leave uncovered many areas that should receive protection. Representative Franks bill, the Children Privacy Protection and Parental Empowerment Act also provides important safeguards.
Establish a Privacy Agency
In 1973 the Department of Health, Education and Welfare established a special panel to study privacy issues arising from the growing use of automated date processing equipment. That report led to the development and passage of the Privacy Act of 1974, perhaps the most important privacy law in our country. But that report also made clear, as have subsequent reports, that the cornerstone of an effective federal policy is a permanent privacy agency.
It is critical today that a privacy agency be established. We simply do not have the expertise, commitment, or understanding in the federal necessary to develop the policies necessary to address the enormous challenges that we are facing. Many of the decisions that are made with significant consequences for privacy protection lack adequate representation of privacy concerns.
In countries across the world, efforts are underway to address these privacy concerns. The European Union is moving forward on the implementation of extensive privacy directive that will establish legal rights for all citizens in the European Union countries. Non-EU countries, including Japan and Canada, are pursuing comprehensive privacy polices. Techniques for anonymity are being promoted in Germany, the Netherlands and elsewhere. Strong medical privacy legislation is in place in New Zealand.
In the United States, even with the efforts of the Federal Trade Commission, there is little sense that we are making progress. Privacy concerns are rising. The public is not persuaded by the current policy. BusinessWeek put it well in an editorial earlier this month:
Time is running out for the Net community. The public does not trust its promises for self-regulation to ensure privacy. The polls show that people don't believe that these voluntary standards are working. Any spot check of Web sites shows that few make any serious effort to protect privacy. It's no wonder that the public wants the government to step in immediately and pass laws on how personal information can be collected and used. Even Silicon Valley libertarians who believed in voluntary standards for years are no longer so sure.
As the economy shifts increasingly from an industrial to an information base, an individual's private data take on an economic utility unknown in the past. So, too, does a person's economic behavior in the electronic realm. Future growth depends on the security of that data and the comfort level for that behavior. Both civil society and economic growth depend increasingly on privacy.
The United States has long been a beacon of individual liberty and a champion of individual rights. Our greatest challenge today is to carry forward that tradition into the information age. For Internet users today and into the future, that will mean protecting the right of privacy.
P. Agre and M. Rotenberg, eds., Technology and Privacy: The New Landscape (MIT Press 1997)
J. Cohen, "A Right to Read Anonymously: A Closer Look at Copyright Management in Cyberspace," U.Conn.L.Rev. (1996)
W. Diffie and S. Landau, Privacy on the Line: The Politics of Wiretapping and Encryption (MIT Press 1997)
S. Friewald, "Uncertain Privacy: Communication Attributes After The Digital Telephony," 69 S. Cal. L. Rev. 949 (1996)
International Working Group on Data Protection, Data Protection and Privacy on the Internet, Data Protection and Privacy on the Internet (1996) [http://www.datenschutz-berlin.de/diskus/13_15.htm]
National Information Infrastructure Task Force, Information Policy Committee, "Options for Promoting Privacy on the National Information Infrastructure" (1997)
Organization for Economic Cooperation and Development, Guidelines on the Protection of Privacy and Transborder Flows of Personal Data (1980) [http://www.oecd.org/dsti/sti/it/secur/prod/PRIV-EN.HTM]
Organization for Economic Cooperation and Development, Cryptography Policy Guidelines (1997) [http://www.oecd.org/dsti/iccp/crypto_e.html]
P. Regan, Legislating Privacy: Technology, Social Values, and Public Policy (University of North Carolina Press 1995)
M. Rotenberg, "Communications Privacy: Implications for Network Design," Communications of the ACM (1995)
M. Rotenberg, "Data Protection in the United States -- A Rising Tide?" The Computer Law and Security Report 38-40 (January-February 1998)
M. Rotenberg, "In Support of a Privacy Protection Agency in the United States," Government Information Quarterly (Winter 1991)
P. Schwartz and J. Reidenberg, Data Privacy Law (Michie 1996)
B. Schneier and D. Banisar, The Electronic Privacy Papers (John Wiley 1997)