Saturday, April 14, 2007

Who’s Watching the F.B.I.?


Published: April 15, 2007


In “A Midsummer Night’s Sex Comedy,” Woody Allen shows up at Mia Farrow’s window on a flying bicycle and urges her to hop on. “Andrew, we’ll get killed,” she protests. “Trust me,” he replies, “it’s me, Andrew.” She looks skeptical, and he tries again. “Trust me anyhow.”

In the latest and most serious post-9/11 civil-liberties abuse to emerge from Washington, the Bush administration’s “Trust me anyhow” defense has finally collapsed. The scandal involves “national-security letters,” which the F.B.I. has secretly used to scrutinize the financial data, travel records and telephone logs of thousands of U.S. citizens and residents. In March, a report by the inspector general of the Justice Department described “widespread and serious misuse” of national-security letters after the U.S.A. Patriot Act of 2001 significantly expanded the F.B.I.’s authority to issue them: between 2003 and 2005, he concluded, the F.B.I. issued more than 140,000 national-security letters, many involving people with no obvious connections to terrorism. The Bush administration was fortunate that, shortly after the F.B.I. scandal broke, the tempest over the Justice Department’s firing of prosecutors bumped it off the front page.

National-security letters are especially susceptible to abuse because they’re not subject to independent review by a judge or magistrate and because the recipients are forbidden to discuss them. In an op-ed article published anonymously last month in The Washington Post, the president of a small Internet-access business described the “stressful and surreal” experience of receiving a national-security letter. Under threat of criminal prosecution, he wrote, he was forbidden to discuss any aspect of the case with his colleagues, his family, his girlfriend or the client whose data he had been ordered to reveal.

National-security letters were authorized in 1978 as a narrow exception to federal privacy laws, and their reach was expanded in 1986 to give the F.B.I. easier access to the records of suspected spies. The F.B.I. could issue the letters only if senior officials in Washington had a factual basis for believing that the records pertained to a suspected spy or terrorist. But the Patriot Act diluted these requirements, allowing F.B.I. field agents to issue the orders on their own say-so merely by asserting that they were “relevant” to a terrorism investigation.

Critics warned that these changes would let the F.B.I. collect the personal data of Americans with no clear ties to suspected terrorists, but few predicted the magnitude of the F.B.I.’s incompetence. The inspector general’s report found that the F.B.I. wasn’t even following its own internal guidelines and in some cases had violated federal law. The bureau wasn’t keeping signed copies of all its national-security letters and, as a result, couldn’t properly track the data it got. In the spirit of the Keystone Kops, it didn’t realize when it received data on the wrong person. When an F.B.I. official complained to his superiors, he was ignored.

It is too simple to attribute the F.B.I.’s abuses to plain carelessness. They also reflect the bureau’s struggles to reinvent itself as an agency devoted not merely to prosecuting past crimes but also to preventing future ones. The prevention strategy is based on the idea that the best way to avoid future 9/11’s is to collect information on lots of people who aren’t obviously terrorists and prosecute them for minor crimes before they have an opportunity to blow up buildings. This necessarily encourages dragnet searches of millions of people who turn out to be innocent. But even if F.B.I. agents clear the subject of a national-security letter, they store the information gleaned from it in digital databases that include more than a half-billion records, are not purged for at least 20 years and can be shared with state law-enforcement agencies or with private businesses for data mining.

Because of the amorphous nature of the terrorist threat, the F.B.I. may be right that it needs the power to investigate people who it isn’t sure in advance are terrorist suspects. Predicting who might be a terrorist in the future is much harder than prosecuting known spies who committed crimes in the past. But dragnets have their price.

Fortunately, some in the new Congress have indicated that they intend to revisit the Patriot Act. The broad outlines of the necessary reforms have long been obvious. Congress needs to restore independent review by judges in cases where the Patriot Act eliminated it, ensuring neutral oversight of secret searches. Those searches should be focused on the associates of suspected terrorists, rather than sweeping up any citizen who has information that might possibly be relevant. And Congress should restore a degree of transparency by lifting the gag orders in secret searches after a reasonable period of time.

At a hearing before the House Select Intelligence Committee last month, Justice Department and F.B.I. witnesses, sticking doggedly to their script, said that putting courts in the middle of the process would slow it down. This time, however, the “Trust me anyhow” defense had few takers. “I think self-policing has failed horribly,” said Representative John Tierney, a Massachusetts Democrat. If Congress continues to focus on the F.B.I. abuses, many Americans may come around to the same view. Unlike the National Security Agency’s warrantless surveillance of suspected terrorists abroad, which supposedly involved them, the F.B.I.’s domestic surveillance clearly involves us.

Jeffrey Rosen, a frequent contributor, is the author most recently of “The Supreme Court: The Personalities and Rivalries That Defined America.”