Probably the more important outcome of this episode would be guidelines on agent recruitment that the DCI issued after the Bamaca affair. These strengthened strictures on employing rights violators as agents. Much has been made of the restrictions, especially since the 9/11 attacks, as having tied CIA’s hands. But Director Deutch recognized at the time the need to employ spies of poor reputation; his regulations essentially provided that shady characters had to be approved by headquarters. Since Langley already routinely passed on recruitments, the regulations add only a shade of human rights emphasis. In 2002 the CIA issued an official announcement revealing that it had never, in fact, been prevented from hiring anyone as a result of the 1995 regulations.

  These events taught operations officers that the Seventh Floor did not support the troops and accentuated the trend toward headquarters absorbing all important decisions. Melissa Mahle makes the relevant point that the burgeoning sophistication of the computer and the internet, which made possible instant communication between Langley and stations, added to the temptation to micromanage the field. John Deutch, by trade a mathematician and the innovator of cybernetic theory, might have been pleased, but DO people were not. Mahle records being encouraged to take out liability insurance. A million dollars’ worth became a baseline figure. Frustration rose steadily, so high by 1999 that eight case officers sued the agency, charging that the CIA routinely discouraged their resort to lawyers in disciplinary matters and then failed to provide data necessary to a defense. “The field,” Mahle writes, “became irrelevant.” It is a further irony of history that the same lawyers who were warhorses of the secret war at the dawn of the CIA have become contributors to its risk-aversion.

  Deutch cast his approach as “risk management,” a term that could have come right out of cybernetic theory but that really meant arranging approvals so as to spread blame among the layers of the high command. It was not long before observers and former CIA officers began decrying the “risk averse” agency.

  On straight management issues, Deutch also ruffled feathers. The director wanted his own man for DDO. At the time Jack Devine served as acting deputy director, the result of his post as associate, to which Woolsey had promoted him. The DO’s favorite, Jack G. Downing, cut from the classic spy mold, was the only officer who had headed the stations in both Moscow and Beijing, and had been East Asia baron too. Devine, the Latin America expert, could have been DDO himself. Deutch passed over both to select David Cohen, a career CIA analyst whose only experience in operations had been a brief tour as head of the Office of Domestic Resources, not a cutting-edge DO position and a decade in the past to boot. Deutch sent Devine to London as station chief. Downing retired. Cohen proved inept at guarding DO interests.

  Deutch brought Nora Slatkin with him from the Pentagon as executive director, the agency’s senior management assistant. Many at CIA considered Slatkin a neophyte on intelligence and completely beholden to the top boss. Slatkin sightings and stories proliferated at Langley faster than the old Soviet Union had built nuclear weapons. The DO employed a special assistant just to keep track of her movements within the building and warn of her appearance. Among the bloating that went on during this period, Slatkin replaced the three-person unit supporting the executive director on covert action approvals with a staff of a dozen, all valuable DO officers then unavailable to the field.

  The Covert Action Review Group became just one of many reforms, for Deutch pulled out the stops. He wanted officer participation, and many became involved. An agency that prided itself on being the least hidebound in Washington became a maze of hoops and ladders. One promise he delivered on would be improving the CIA relationship with Congress. Here he had a major assist from DDCI George J. Tenet, who came from the NSC staff where he had been top man for intelligence. Before that Tenet had spent a decade on Capitol Hill, mostly on the staff of the Senate intelligence committee, rising to become its director.

  A couple of years earlier Woolsey had favored shutting CIA stations in Africa. Deutch talked of closing down much of the DO empire, preserving just major stations, in favor of technical collection that would presumably make up the gaps thus created. That changed. In April 1996 Deutch refused to promise Senate overseers that CIA would refrain from using journalists. By summer he spoke in Clinton’s councils in favor of restructuring the remaining elements of the Iraqi covert project. In September Deutch began saying in speeches that he intended to increase CIA personnel assigned abroad and recruit more agents, and that he favored more covert operations. That November Deutch flew to Los Angeles for an unprecedented appearance at a town meeting to defend the DO against charges the CIA had colluded with drug traffickers in California and Central America. Having famously said that CIA officers were not as competent as the staffers he’d known at the Pentagon, in a departure interview with the Washington Post Deutch said, “I fired a lot of people. I certainly could have fired more . . . after a while, you have to ask, what are you accomplishing by doing it.” At the end, Deutch felt, he and the denizens of Langley had begun singing from the same sheet of music.

  Amid all the angst, CIA operations had to go on. The most important remained Bosnia. Agency planners discovered they had just four officers who spoke even passable Serbo-Croatian. H. K. Roy was one of them. Roy had seen the disintegration of Yugoslavia from the CIA station in Belgrade, had covered Kosovo, and had done an earlier mission in Bosnia. He had learned of the Lake-Galbraith arms ploy only a few months after it took place. In mid-1995 the DO selected him for Sarajevo as its first station chief. The “station,” however, consisted of just one communications officer, a man who had been in on the chaos of The Dish, perhaps fitting preparation for Sarajevo. Roy helped pinpoint Serbian gun positions that would be destroyed in Operation Deliberate Force, and he reported on the massacre at Srebrenica. Then Roy discovered the downside of the Iranian arms traffic—the Iranians controlled Bosnia’s interior ministry. He learned Iranian agents planned to kidnap him. Suddenly exfiltration became necessary.

  At one point the situation grew so taut that Roy made his radioman promise to shoot him if they were about to be captured. In token to the complexities of the Deutch-era CIA, Roy had had to sign a waiver indemnifying the CIA in case he were disfigured on this mission (unlike, for example, Dick Holm in the Congo). Both Roy and his assistant realized that, if the eventuality arose, his comrade would have to convince agency lawyers that killing the station chief had been in Roy’s best interest.

  The Clinton administration eventually demanded—and received—the dismissal of the offending pro-Iranian minister as a condition for the military aid and training program it sponsored for Bosnia-Herzegovina.

  Roy’s was not the only exfiltration. Admiral Blair outraged the DO when he talked of agency sources to reporters, compromising several agents who then had to be moved to safety. Some CIA people also had to be redeployed. The cost reportedly came to $10 million. It is also said that a woman officer died in the line of duty in Bosnia.

  Center stage now went to Support for Military Operations, the notorious SMO. Contingency operations, like Somalia, Haiti, and now Bosnia, were the major form of U.S. military activity. Admiral Blair presided over formation of a National Intelligence Support Team in tandem with Clinton’s dispatch of an expeditionary force. In January 1996 the mission went into high gear. It included CIA operatives and analysts, military folk from the Defense Human Intelligence Service and DIA, NSA cryptologists, technologists, and communicators. Support included a U-2 detachment flying from southern France, early use of computer terrain-modeling, and the first use of the unmanned aerial vehicle known as Predator. The initial deployment was handled by a CIA team of fifteen under a junior officer. By all accounts the Predator did splendidly. When the military tried the same thing, they needed a unit of two hundred airmen headed by a colonel. Two of their eight Predators were lost within weeks.

  The intelligence support team worked in Bosnia for the long term. Some variant of it continues there today. New intelligence teams became necessary in 1999 when Clinton moved on to confront the Serbian Republic directly in the Kosovo war. The highlight there, corrupted location data that led to the U.S. Air Force bombing the Chinese embassy in Belgrade, naturally became a matter of dispute. At least the Directorate for Operations rested easy—blame went to other segments of the intelligence community.

  In November 1996 John Deutch resigned, following the tradition of allowing an incoming president—Bill Clinton returning for a second term—to choose his senior officials (Deutch held cabinet rank). He hoped to go to the Pentagon as secretary of defense. Instead Deutch received no post at all. In his memoir Clinton says that he “hated” to lose Deutch at CIA, but no other evidence exists to substantiate this. By then so much bad blood spotted the rafters that when computers at Deutch’s home were found to contain unprotected classified information, CIA security officers were ready to throw the book at him. Deutch lost his CIA and (for a time) Pentagon security clearances, finally avoiding criminal prosecution only because of a presidential pardon. Nora Slatkin and five others were reprimanded for their handling of the original inquiries.

  Capitol Hill watched all this closely. In fact the Senate Select Committee on Intelligence held four hearings before October 1996 just on the question of intelligence for intervention in Bosnia. That would be only one SSCI probe. Congress worked on intelligence reform legislation it put into CIA budget authorizations. Other areas of inquiry included examining the CIA role in Gulf War intelligence before the exposure of U.S. troops to potential toxic chemicals, the CIA’s use of journalists and others, allegations of a CIA connection to contra drug trafficking, the claims upon the U.S. government by former South Vietnamese commandos who had fought for the CIA, the Bamaca case in Guatemala, intelligence for law enforcement, information security, the inspector general’s office, organized crime in Russia, airborne and satellite reconnaissance, and more. The Senate intelligence committee also investigated terrorism issues, including the June 1996 bombing of the Khobar barracks in Saudi Arabia.

  Counterintelligence continued to be a concern. Not only did the Aldrich Ames case still make waves, in November 1996, on Deutch’s watch, the FBI arrested a DO counterintelligence officer, Harold Nicolson, who pled guilty to spying for the Russians.

  In the time Dick Helms sat at the desk on the Seventh Floor, months’ worth of CIA contacts with Congress could be listed in double-space typescript on just one or a few sheets of paper. That had not been the whole story, of course; there were a certain number of chats on the phone or at Washington parties. But by the 1990s contacts with Congress were running considerably higher than a thousand a year—and those were merely the official ones. Langley seriously discussed creating a regular intelligence periodical for Congress along the lines of the President’s Daily Brief and similar reports.

  Between January 1995 and September 1996 the SSCI received no fewer than 436 notifications of significant intelligence activities, not because so many covert operations were under way but because notification requirements expanded to encompass extremely minute actions. That too was an index of the rise of the congressional intelligence juggernaut. Conditions for notification, especially the definition of “current” notification, were also still controversial. In fact that issue figured in the most important SSCI inquiry of 1996, about Bosnian arms.

  By far the key oversight action of this period with respect to covert operations was a series of inquiries into the Bosnian arms traffic. On April 5, 1996, the Los Angeles Times began publishing a succession of articles revealing the Lake-Galbraith maneuvers with their “no instruction” subterfuge. No fewer than six congressional committees took this up. In the House of Representatives, which had shifted to Republican control in the previous election, the incident offered an opportunity to score political points. (Some in the House probably saw a chance at payback for Iran-Contra.) The House International Relations Committee created a subcommittee and spent a million dollars on its inquiry. The SSCI developed identical information and reached the same conclusions in an investigation conducted in the normal order of its business. These were the most important of the congressional probes into the Bosnia affair, yet aside from generating heat and a huge report there was little to show for the effort.

  For present purposes the Senate intelligence committee inquiry is the more relevant because it was completed and the report issued on November 7, 1996—still on Deutch’s watch—and because President Clinton shortly thereafter nominated Anthony Lake as CIA director in his place. The Senate would have to confirm Lake, and the intelligence committee had jurisdiction. Here lay the ground on which the latest round of the legislative-executive struggle for control over U.S. intelligence was fought out.

  Lurid press accounts pictured the Clinton White House as purposefully evading a UN embargo, in league with nasty Iranian ayatollahs. The Congress inevitably wanted to sink its teeth into that. Never mind that the Senate had pressured Clinton to lift the embargo, passing some of its legislation at the very moment of the Lake-Galbraith gambit, or that a few months later legislators actually moved to prohibit Clinton from enforcing it. The SSCI had several public hearings to question Peter Galbraith, Charles E. Redman, Strobe Talbott, and Richard Holbrooke; four more closed sessions were held plus six “informal” meetings. Witnesses included John Deutch, Jim Woolsey, the chairman of the JCS, the secretary of defense, and State and Defense department lawyers. Tony Lake appeared before an informal meeting, putting another crack in the dike presidents traditionally use to wall off their national security advisers from congressional scrutiny. Anthony S. Harrington of the IOB presented the board’s conclusions to the senators as well.

  Everyone insisted the Bosnia initiative had been a diplomatic maneuver, that impressions of a covert operation were mistaken, and that planning had never led to operations. The senators eventually accepted this view. The SSCI report argued that the administration would have been better off to inform participants of what had been afoot, and the report especially singled out lack of notification to Congress, which it deemed unacceptable. That led right back to the debate over “current” reporting of covert activity that had raged so long and so bitterly. The Senate committee also resurrected the issue of “third country” operations and notification responsibilities in those cases. Iran and others helping arm the Bosnian Muslims obviously fell under “third country” provisions, and the SSCI report reminded all that oversight in this area had yet to be perfected.

  President Clinton’s nomination of Tony Lake for CIA director afforded the Senate an instant opportunity to press its concerns and a position of advantage. John Deutch introduced Lake to CIA personnel and gave him a bear hug at an event in The Bubble, but that counted for little on the Hill. Not surprisingly, in the phone calls and informal meetings a nominee typically has before appearing—horse shows as it were—Lake told senators that the administration had erred in not informing Congress of the Bosnia ploy. He promised better. Lake also made the proper genuflections to legislative oversight, despite the secrecy he clearly had favored as national security adviser.

  The senators had other concerns about Anthony Lake too. Some of them responded to Langley’s doubts, given that Lake seemed to have had little time for CIA during his White House years. Lake mobilized support to counter such complaints. An anonymous “senior intelligence officer”—likely John Deutch—told the New York Times: “My experience with Tony Lake regarding secret operations, clandestine operations, and the clandestine service has been absolutely outstanding. I could not ask for more support. With respect to covert action programs, he’s helped us focus them on the right things. That’s had a dramatic and positive effect on my officer corps.” This would have been news to Bob Baer, whose anti-Saddam coup Lake canceled the day before its execution.

  At first the conventional wisdom held that Lake would be confirmed. Then questions arose over his stock portfolio, held a couple of years into the administration; then allegations that Lake had prevented the briefing of Congress on charges of secret Chinese campaign contributions in 1996; then others. Lake’s promises to put his stocks in a blind trust had no impact.

  The real issue, cloaked by the palaver, remained the legislative-executive balance of power over intelligence, exemplified by the Bosnia gambit. That became even clearer when SSCI chairman Richard Shelby, a Republican from Alabama, finally permitted confirmation hearings to begin—after three months had passed. Anthony Lake spoke of unvarnished intelligence. But senators’ questions concentrated on Bosnia, the absent China briefing, Lake’s belief that leakers were the same as spies, his opinion as to whether Alger Hiss had been a Communist—not much about the business of the CIA. While hitting Lake for suppressing data, Senator Shelby demanded and obtained access to the FBI background investigation for the Lake appointment—another indication of the balance of power.

  When Shelby wanted another delay, supposedly to permit others to review the report, Lake threw up his hands. Bill Clinton recounts, “He was worn down after working seventy- and eighty-hour weeks for four years. And he didn’t want to risk hurting the CIA with further delays. If it had been up to me, I would have carried on the fight for a year if that’s what it took. . . . But I could see Tony had had enough.” The former president comments, “I still regret the raw deal handed to Lake.” Tony Lake withdrew his nomination with a testy letter to President Clinton, promptly released by the White House, calling the SSCI process a “political circus.”

  After the Lake episode Bill Clinton understood that the next director of central intelligence had to be someone who could please Congress and obtain confirmation from the Republican majority in the Senate. The president wanted someone acceptable to him, and it would not hurt at Langley if the candidate were in with, or at least known to, CIA officers. Several agency officers met the third criteria but not the others. The acting DCI, Deutch’s deputy, fit every one. George Tenet had had two years to learn the ropes at Langley, and now he looked ready to step up. Tenet had got on well with Clinton while on the NSC staff, and he had extensive contacts on the Hill after years on the SSCI staff preceding that. A New York boy but close to the working class—his father owned a Greek deli—the sophistication Tenet had won the appreciation of the president, whose own rise from hardscrabble Arkansan to Oxford scholar held many similarities. President Clinton quickly named Tenet for DCI, and the nominee just as quickly won Senate confirmation.