In March, 2003, the FBI arrested a Chinese-American businesswoman and Republican fundraiser, alleging that she had passed a frighteningly broad range of American intelligence secrets to the People's Republic of China (PRC). For two decades, Katrina Leung had been a paid bureau informant, supplying information on Chinese intelligence operations in America. She'd also been sleeping with two senior FBI agents--one of whom was her so-called "handler"--for the better part of those two decades. It was alleged that she had transmitted what she learned about American counterintelligence from her lovers to Beijing and sent Beijing's disinformation back through the FBI. The story was sordid, embarrassing, and, worse than that, quite grave: Intelligence sources told The Washington Post that Leung had single-handedly compromised 20 years of American counter-intelligence work against the PRC.
Democrats, who in 1997 weathered endless--and ultimately unproven--accusations of selling political favors or national security secrets for PRC money, can take a measure of satisfaction from this unlikely coda: The only bonafide Chinese spy so far turns out to have been not only a Republican, but a well-connected GOP fundraiser. And not just any Republican fundraiser, but one who happened to be sleeping with one of the lead FBI agents investigating Democratic fundraising.
t's bad enough that Leung was able to seduce two FBI agents. But her longtime handler and lover, James Smith, was in possession of information covering a wide range of investigations and operations aimed at the PRC. Since Smith had access to so much, and Leung had access to what Smith had (copying and returning documents from his briefcase before he noticed their absence), her treachery touched everything: the 1997 campaign finance scandal, the investigation of Wen Ho Lee (the Chinese scientist at Los Alamos who was once suspected of selling nuclear secrets to Beijing), investigations of spies at the Lawrence Livermore National Laboratory, and much more. "They lost everything," one hawkish D.C.-based China watcher told me. "It's not how big a fish she is; it's how much damage did she do to the system over 20 years. She totally wrecked it." The real lesson that the Katrina Leung case teaches is one that the FBI and the Republicans, who became its most aggressive patrons during the 1990s, have spent almost two decades ignoring: The repeated failure of the FBI to adopt basic counterintelligence tactics has left it wide open to moles and spies.
From time to time, every spy agency falls victim to a mole, a traitor, or a double agent. It's in the nature of the enterprise, since each such institution constantly attempts to penetrate the secrets of almost every other intelligence service. But because intelligence professionals know that it is extremely difficult to guard against every compromise of an agency's secrets, they are supposed to structure their outfits in such a way as to minimize the damage when the inevitable breach occurs. The best way to do that is through what intelligence professionals call "compartmentation"--designing the organization like a honeycomb, with individual parts sealed off from the rest as much as possible, and distributing information within the organization only on a "need to know" basis. There's always a tension between the needs for compartmentation and information sharing. But without effective compartmentation, a single, well-placed mole can trigger an intelligence leak of catastrophic proportions. Poor compartmentation also makes finding the culprit almost impossible.
If the Leung scandal were a one-time goof, it might not be so outrageous. But it's not. The problems it exposed bear striking similarities to those revealed in the investigations into the Soviet-controlled American spies Aldrich Ames and Robert Hanssen--problems of information security about which the bureau had been repeatedly warned, but had just as often failed to address. This is no insignificant bureaucratic rigidity. Some of the country's most important national security secrets over the last 20 years have been exposed to our two biggest adversaries, and finding the culprits has been long delayed because of the bureau's failure to effectively implement this most basic principle of intelligence work. Despite no fewer than five very public warnings, Washington has been chronically unwilling to fix it.
These repeated, dangerous failures at the FBI have both administrative and political sources. Bureaucratically, the agency is being asked to undertake two incompatible responsibilities: law enforcement and intelligence work. Though the two activities are related and overlapping, the skills, strategies, and tactics needed for each are profoundly different. The skills needed for law enforcement--a clubby culture of sharing information among agents--often means disaster in intelligence work. The latest debacle is proof that the bureau has never, and can never, overcome this built-in conflict in its mission. Only changing the FBI's mission can solve the problem. But only politicians can change the bureau's mission, and that's the second, more disturbing source of the problem. For the bureau's serial failures have been revealed at a time when Republicans have been tightening their hold on power in Washington--including on the congressional committees that oversee the FBI. Equally important, it has been during this period that the GOP has chosen to act as the FBI's protector, encouraging its investigations of the Buddhist-temple affair and other "scandals" which hurt the Democrats, while shielding the FBI from tough but necessary reforms that might have stopped the real damage done by spies like Leung.
Many of these problems festered during the Clinton years. And thus some measure of the responsibility must, by definition, fall on the former president's shoulders. But his ability to impose reforms on the FBI was stymied by numerous politically inspired investigations which would have made any pressure from the White House appear politically motivated. The Republicans decided not to act because, for them, the politics were just too good.
The Hoover Vacuum
The FBI has always been deeply resistant to reform. During the 1950s and 1960s, under J. Edgar Hoover, the agency routinely spied on American citizens--eventually surveilling and compiling dossiers on a good bulk of the American political, entertainment, and literary establishments. Congressional oversight was virtually non-existent; and the FBI mostly did as it pleased. But after Watergate, a prevailing climate for reform did develop, and the Church Committee report reined in the FBI, bringing about a sharp limitation in its powers of domestic surveillance--proof, say law enforcement experts on all sides of the political spectrum, that the FBI can be reformed, when there is the political will to do it.
But there was one crucial, structural problem the Church Committee didn't fix: the conflict between law-enforcement and intelligence. The key differences between police work and intelligence work are rooted in their divergent attitudes toward suspicion and information-sharing. And the law-enforcement attitude suffuses the culture of the FBI, because the bureau has always been dominated by its agents--their specific designation, with intentional grandiloquence, is "special agent." Once a recruit attains this rank he or she is a member of a brotherhood. And the bureau has been notoriously unwilling to suspect its own of such serious transgressions as treason, espionage, or even serious malfeasance.
That is precisely the opposite of what is required of an intelligence agency. Though esprit de corps is undoubtedly important, intelligence agencies are supposed to operate on a principle of suspicion. No one is ever truly above doubt. The bureau has historically prided itself on a culture of gun-toting and case-cracking. The more sedentary, information-intensive side of law-enforcement--increasingly important to law enforcement today, but always a mainstay of intelligence work--has long been shortchanged. What the FBI calls "intelligence analysts" are more like researchers assigned to special agents running cases--a low-status position indeed within that proud, macho, knock-down-the-door organization. As the bureau jibe has it, real men don't type.
Finally, law enforcement places a far higher priority on the fast and easy exchange of information within the police brotherhood. The FBI "is a culture that prized sharing information, not compartmenting it," says Gregory Treverton, a former Church Committee staffer who served as vice chair of the National Intelligence Council in the early 1990s. After all, that's how cases get cracked, but not how traitors get found--it just makes traitors' work easier. And it has simply proven impossible to maintain one modus operandi on one front of the organization and a different one on the other. This is especially so since there will always be more law-enforcement professionals in the FBI than intelligence and counter-intelligence operatives. And, consequently, in institutional terms, the former will always dominate the bureau's leadership and set the organization's tone.
Mountains and Mole Spills
In 1985 and 1986, the American intelligence community had a serious problem on its hands: a series of American informants in Russia run by both the CIA and the FBI disappeared, only to turn up dead or in prison. Though most people think of the CIA as the agency that handles overseas espionage, the FBI has also long conducted espionage against foreign countries. And it is the FBI's responsibility, as the nation's chief law enforcement agency, to conduct criminal investigations of all breaches of American intelligence, whether they occur at the FBI, the CIA, or elsewhere in the intelligence community. So, after the disappearance of the Soviet informants, the FBI organized a task force to discover how U.S. intelligence assets in the Soviet Union had been compromised. The evidence strongly indicated that at least one spy was well-placed within the U.S. intelligence community. By 1988, the bureau's counterespionage branch had concluded that a mole was the only credible answer. But, where?
Unfortunately, the FBI's investigation could hardly get off the ground. The FBI was so poorly compartmented that just in the Washington field office hundreds of people, from senior staff down to secretaries, knew information that made its way into Soviet hands. That lack of compartmentation not only ensured that a mole would have had access to a very broad range of classified information, it also greatly complicated the task of looking for and detecting such a mole. For this and other reasons, the FBI's investigations--of the CIA and of itself--eventually ground to a halt. Nor did the bureau, or its overseers on the Hill, make any attempt to remedy the underlying problem of lax compartmentation--both in the Washington field office and elsewhere within the bureau--which its own investigation had uncovered.
The investigation was moribund until 1991, when the CIA uncovered financial information which suggested the mole might be Aldrich Ames, once head of the CIA counterintelligence office in Moscow, who had been living far beyond his means. The CIA called in the FBI, who looked through the bank records and eventually arrested Ames as a spy. Ames pled guilty; over the course of eight years, he'd sold out dozens of American informants, including General Dmitri Polyakov of Soviet military intelligence, widely considered one of the CIA's most useful agents ever. Ames was paid a total of $4.6 million by his Moscow handlers for his espionage, and is currently serving a life sentence in federal prison.
In 1996, two years after Ames's arrest, Justice Department Inspector General Michael Bromwich set out to discover why it had taken some eight years for the FBI to arrest Ames when simple bank statements, open high living, and other readily available evidence seemed to point so clearly to his guilt. Bromwich found that from the start the bureau leadership had turned the investigation over to relatively low-level agents, given the investigation indifferent attention, and allowed it to break down entirely at several points without coming up with any reason for the disasters.
But there was more. Bromwich found that the FBI had not only been seriously deficient in scrutinizing its own agents, it had also fallen down at even some of the most basic, established methods of counterintelligence work. Bromwich's investigators found that the bureau's Counter-Intelligence Division practiced little of the compartmentation which is the sine qua non of serious intelligence work. According to Bromwich's report, "as many as 250 FBI employees at the FBI's Washington Field Office alone likely had knowledge" of the key information in question. The FBI wasn't responsible for Ames's treachery--that was the fault of the CIA--but problems at the FBI prevented the bureau from uncovering it for years.
The Bromwich Report was a crucial, and very public, criticism of the ways in which the FBI was run, but neither the FBI itself nor its political overseers on Capitol Hill took action. The politics of the time dictated that neither Democrats nor Republicans were willing to censure the FBI, a result of the careful political position that FBI Director Louis Freeh had carved out for himself since his appointment four years earlier. In 1993, then-President Clinton had promoted Freeh, a judge, one-time FBI agent, and registered Republican, to replace the FBI chief Clinton had inherited from the Bush administration, William Sessions. Freeh stepped into a treacherous political climate. Republicans were incensed at the FBI for their ill-conceived--and deadly--1992 raid on the Ruby Ridge, Idaho, cabin of armed white separatist Randy Weaver. Even worse for Republican critics of the FBI was the mishandled 51-day federal standoff with David Koresh and his Branch Davidian cult in Waco, Texas, which ended with 80 dead when the cultists burnt their own compound to the ground. Seventy-six cultists were killed, along with the four federal agents killed in the original raid. The FBI was also facing a different, more critical Republican party in the mid-1990s, filled with figures like Helen Chenowith, the then-Idaho congresswoman given to dark mutterings about black helicopters. Then, in 1994, Republicans took over Congress, so all of that built-up hostility to the agency now sat in the seats of power.
Facing this potential onslaught, Freeh made a tacit arrangement with the new Republican barons on the hill, as David Plotz of Slate and others have written. Freeh would focus on multiple investigations of his nominal bosses in the Clinton administration--Whitewater, Henry Cisneros, Mike Espy, Vince Foster--in exchange for a free pass on his and the bureau's many failings. That left problems in counter-intelligence free of either internal or congressional scrutiny. If Clinton administration officials were alarmed about the FBI's compartmentation problems and had plans to fix it--and it's not clear that they were--there was little they could do because of the Republican power on the Hill. Any attempt to rein in the bureau would be seen as an effort to stymie those investigations. In that climate of malign neglect, the bureau's ills were allowed to fester.
The failure of the FBI or Congress to act on the recommendations of the Bromwich Report enabled the career of yet another damaging spy, FBI counterintelligence agent Robert Hanssen, who'd worked as deputy director of the bureau's office of Soviet counterintelligence and FBI liaison to the State Department. He spent two decades pinpointing American informants for Soviet counterintelligence, in exchange for diamonds and cash. By the time he was finally caught, in March, 2001, he'd helped Russian intelligence capture more than 50 American informants.
But the problems which allowed Hanssen to elude capture for so long must have been familiar to FBI leadership. Hanssen was able to supply secrets from all sorts of FBI intelligence and counterintelligence operations because the FBI had still not erected any effective barriers to cross-department information sharing. Sitting at his own computer terminal, Hanssen had unrestricted and untracked access to the whole FBI data system--including details on the mole hunt operation which ended up capturing him. Given his high rank in the bureau's counterintelligence operations against the Soviet Union, even very tight compartmentation would have allowed him access to a good deal of information. But laxity in compartmentation allowed him free access to all manner of information his proper duties gave him no need to see. That flaw delayed his eventual capture for years. And his efforts to gain access to such information should themselves have raised immediate red flags. If the FBI had taken to heart the recommendations of the Bromwich Report in 1997, it is quite likely Hanssen's final phase of spying never would have occurred. But even after the FBI caught Hanssen, neither the White House nor the Republican leadership in Congress pressed the FBI into making belated reforms.
Then came September 11, and with it sudden pressure from the media, the public, and politicians to find out why the intelligence community hadn't anticipated the strikes. Attention quickly focused on the failure of different branches of the intelligence community to share information. If FBI agents in Minnesota had reports of suspicious Arabs taking flying lessons, and so did agents in Florida, why hadn't that information been compiled centrally? The narrative that developed from the September 11 investigations held that the intelligence community needed to break down barriers and share information more readily and rapidly.
But that impulse heightened an existing tension. The lesson from September 11 was that the intelligence community needed to do a better job of sharing information. But the lesson from the Ames, Hanssen, and Leung scandals was that the FBI needed to establish serious need-to-know criteria. Last year, a joint working group of the Senate and House intelligence committees, sifting through reform proposals that had long circulated among intelligence experts in Washington, suggested one possible way to resolve the conflict: Break up the FBI, by hiving off a new domestic counter-intelligence agency modeled on Britain's MI5. The new agency would absorb the FBI's domestic spook-hunting responsibilities, and would be responsible for nothing else. That would permit the FBI to do what it does best--law enforcement--without tearing the fabric of counterintelligence security. A congressional advisory commission headed by former Virginia Gov. James Gilmore endorsed the recommendation. The reform has also been endorsed or entertained by Democratic presidential candidate John Edwards, Republican Sen. Richard Shelby, and numerous intelligence professionals from all points of the political spectrum.
But the Bush administration and the GOP-led Congress eventually demurred. They created the Homeland Security Department under pressure from Democrats, but did little to challenge the bureau's fundamental structural problems. Even in its embarrassed and demoralized state, breaking up the bureau would have required immense political capital. But the year after September 11 was the moment of opportunity for true reform. And probably no president has ever had so much political capital on intelligence and national security matters as George W. Bush did in those crucial months.
While the issue of breaking up the FBI was fading from the front pages in the first months of 2003, Katrina Leung continued to exploit the lack of compartmentation and counter-intelligence procedures at the FBI (a problem which would have at least been on the way to being solved by creating a new domestic intelligence agency) to give crucial American intelligence to the Chinese.
The FBI and its congressional overseers have been given no fewer than five dramatic indications that the bureau has serious deficiencies as an intelligence agency: the Ames, Hanssen, and Leung scandals, each of which stemmed from the same basic problem--poor counterintelligence measures, particularly lax compartmentation--and the Bromwich Report and Gilmore Commission, which put elected officials and the public on no uncertain notice that reform was necessary. But to each of these five challenges, the FBI and Congress failed to respond--and in each case, their failure to act enabled further intelligence failures.
This wasn't simple benign neglect. Some of the nation's most tightly-held and vital secrets were turned over to adversary states. That's the kind of failure that usually drives Republicans around the bend, and for good reason. The mere suggestion that this might have occurred in the Democratic Chinese fundraising scandal aroused paroxysms of GOP outrage: from the wildly overheated Cox Commission Report, to limitless hours of talk radio chatter, to Republican Sen. Fred Thompson's hearings, all pursuing a line of allegation--that Red Chinese money had bought favors in the American political system--that proved unfounded.
Now we have an actual Chinese spy--charged, though not convicted--who by all indications was funneling money into U.S. campaigns. Her treachery is an intelligence failure that comes on the heels of others tied to similar shortcomings at the FBI, and one in which vital secrets were given to a power, China, which these same Republicans were saying two years ago posed the greatest threat to the United States. And yet we've not had one hearing. Not one commission. There's been very little coverage in the press, nor is anyone yakking about it on talk radio.
The Republicans didn't create the problems at the FBI. But they've sat on their hands and put politics ahead of the national interest as the scope of the problem and the cost to national security have become increasingly apparent. Not only have they ignored the problem, they have actively sought to shield the FBI from the one reform that almost everyone agrees would make such breaches of national security secrets far less likely. That's not just politics as usual. It's not even garden-variety political hypocrisy. It's a betrayal of the public trust.