AAAS S&T Policy Forum Explores Privacy in an Era of Powerful Surveillance Tools

This is the age of watching and being watched. Cameras and other sensors have become so small they can be embedded in door knobs and other places people don't notice, according to Daniel Sui, a specialist on surveillance technologies at Texas A&M University who spoke at the recent AAAS Science and Technology Policy Forum in Washington, D.C.

The list of technologies in the new "Viewer Society," as Sui calls it, is expansive. Satellite photos posted to Google Earth have sufficient resolution that one Dutch couple was photographed sunbathing on a rooftop, according to Sui. Aerial photos taken to help monitor coastal erosion in California also provided a glimpse of Barbra Streisand's back yard, prompting her to bring an unsuccessful suit alleging invasion of privacy.

Cell phones and other devices with global positioning satellite (GPS) capabilities allow tracking of users, including teen drivers on the highway or children playing in a neighborhood. Researchers have even mapped real-time changes in the population density of Rome, based on the movement of people using their cell phones.

Infrared sensors can track heat-emitting humans as they move across a landscape. Surveillance cameras on street corners and in stores provide a continual record of every-day activities. Radio frequency identification (RFID) tags, widely used in the identification and tracking of commodities and stock animals, are also available as wearable or implantable identifiers for humans.

All of which raises some important questions: Is secrecy dead in an age of pervasive surveillance technology? Does the definition of privacy need to be redefined? Can science and technology serve both privacy and security interests, allowing researchers to develop still more technologies and data-sifting methods to help prevent terrorism?

Those questions and others were discussed during a 3 May topical session at the AAAS policy forum, which has become the major public meeting in the United States devoted to science and technology policy issues.

While surveillance technologies certainly can raise the specter of an intrusive "Big Brother" government, the picture is more complex, panelists said. In an era of reality television and celebrity mania, the snooping goes both ways, according to K.A. (Kim) Taipale, executive director of the Center for Advanced Studies in Science and Technology Policy in New York. We love to watch others in unguarded moments as long as the camera is turned on them, not on us, Taipale said. Scopophilia or "the love of watching," a concept drawn from the literature of psychoanalysis, has become a cultural, social trait of our times, he said.

In the evolving world of surveillance technology, from the local to the global scale, the boundaries between the private and public realms are being blurred, Sui said. Traditional tenets of privacy, such as "My home is my castle" or "I have a right be left alone," are being broadened to include the right to anonymity and confidentiality, he said.

Taipale said that privacy needs to be "reconceptualized." He argued that "privacy based on isolation is dead…secrecy is over in this world." The best hope, he said, is to manage surveillance systems in ways that allow strict auditing to ensure the collected information is being properly used.

As a practical matter, panelists said, that means finding ways to enable sharing of collected data in a way that protects individual rights while also improving the ability of federal agents to "connect the dots" in anti-terrorism efforts. Tara Lemmey, chief executive officer of LENS Ventures in San Francisco, described some of the work of the Markle Foundation Task Force on National Security in the Information Age (on which she serves).

In a report issued last year, the task force called for an "authorized use" standard for government handling of legally collected information. It would base authorization to view information on how it is going to be used, rather than on the nationality of the subject or the location of collection. The report also proposed an approach to data-sharing that balances the risk of compromising classified information with the risk of failing to share information with those who need it to understand threats to national security.

Lemmey described a system that would break down some of the compartmentalization of information in the intelligence world and allow an FBI agent, for example, to find out more readily whether potentially relevant information exists in a CIA data base. The proposed search system would allow the FBI agent to learn that certain records on a bioterrorism suspect are available. But the agent would be unable to see private or sensitive information without appropriate authorization. The system also would rely on a strict audit system, with tamper-resistant logs, to help ensure compliance with privacy and security rules.

The Markle task force also noted that tools now available or under development can improve information sharing and analysis while protecting privacy. For example, there is an emerging field of "anonymized analytics" in which certain confidential information can be scrambled so that it is no longer readable. The process still allows the information to be compared and matched with other information scrambled in the same manner. If a match occurs, the analyst can request access to the original unscrambled information. The task force said such techniques could be used to determine whether airline passenger records match government watch lists for terrorists while protecting the security and confidentiality of the underlying information.

Another panelist, Don Prosnitz of the Lawrence Livermore National Laboratory, described efforts to develop technologies to prevent terrorism while remaining within Fourth Amendment constraints on unreasonable searches. A Supreme Court decision, Kyllo v. United States, handed down three months before the 9/11 terror attacks, held that the use by police of an infrared sensor to scan a private residence for evidence of marijuana cultivation was unconstitutional. It said use of such a device to learn details unavailable without physical intrusion is presumed to be unreasonable without a warrant.

But, Prosnitz asked, does the decision prevent use of radiation detectors and other sensors to scan city neighborhoods for evidence of weapons such as radiological "dirty" bombs or biological agents? Such broad searches, by their nature, preclude use of warrants.

The Supreme Court held in 1967 (in Katz v. United States) that the Constitution protects people, not places. It said the government may not violate a person's reasonable expectation of privacy. While scientists cannot predict what courts will say in the future about broad use of remote sensors to search for weapons of mass destruction, Prosnitz said there are several criteria that can be used to address traditional constitutional limits. He said a non-intrusive detector that only discloses contraband has the best chance of passing muster.

Prosnitz cited a court case involving a familiar law enforcement tool—the drug-sniffing dog. The court held that the trained dogs were incapable of detecting anything but drugs, which are clearly contraband materials. "A detector that reveals only something that you shouldn't have is okay," he said, and it is not precluded by the Fourth Amendment.

Livermore researchers are building a mechanical "dog," a mass spectrometer that can "sniff" suspect chemicals. The aim is to design a device that, like a dog, discloses only illegal goods. "If you can't interrogate the system to discover anything other than the contraband," Prosnitz said, then the search is probably okay. "I think we can design systems like that," he said. Arms control sensors can sniff a warhead for evidence of nuclear materials without providing details on the warhead design, he noted. Similar information barriers can be designed into a mass spectrometer so that it detects items of interest without revealing other details, he said.

Robert Popp, chief executive officer of National Security Innovations Inc., a consulting business in Boston, said research is crucial in proving the benefits and limitations of various technologies and the means to protect privacy. He formerly worked at the Pentagon and served as deputy of the controversial Total Information Awareness (TIA) program under John Poindexter, National Security Adviser to President Ronald Reagan and a key figure in the Iran-Contra affair.

The TIA was an effort to develop ways to find clues to terrorist activity by sifting through huge numbers of transactions in public, private and government data bases. But the program was derailed by Congress in 2003 after critics raised concerns about mass surveillance of U.S. citizens and potential invasions of privacy. Aspects of the program were funded under a new name, the Terrorism Information Awareness program, and officials have insisted the original program was never designed to create dossiers on citizens. It was, they said, a research effort to improve tools for data collection and sharing.

Popp defended the Pentagon effort, as did Taipale. "There was a misunderstanding about the difference between research and operations," Popp told the AAAS forum. "Opposition to research on the basis that it 'might not work' punishes failure and ultimately stifles bold and creative ideas," he said. The answer to the potential for abuse, he said, is not to prohibit the research on the technologies, which many see as inevitable, but to ensure that the proper checks and balances are in place to protect privacy and civil liberties.

Earl Lane

14 May 2007