Ghost Wars

by Steve Coll

  Miller went to a public park in Kandahar one afternoon and saw some Afghan boys playing. He had thought the Taliban had banned ball games, but now it looked as if maybe some games were okay. As possible gifts Miller had stashed in his truck dozens of neon orange soccer balls and Frisbees. They were leftovers from a Unocal marketing campaign in the United States. All the balls and Frisbees were emblazoned with the Unocal logo. He went back to ask his Taliban hosts if it would be okay to hand out his gifts. They said it would be fine, so he returned to the park and distributed them. Soon the dirt park looked like a neon orange pinball machine with dozens of balls in play and Frisbees sailing through the air.

  A little later, as he tried to schedule a meeting with the Taliban’s assistant foreign minister, Miller shrugged when the minister wondered aloud about when afternoon prayers would be held. A Taliban member at the back of the room, a Caucasian with a long beard and turban, called out in a pungent New York accent: “I think prayer time is at five o’clock.” Miller looked up, startled.

  “Are you an American?”

  He was. His adopted Muslim name was Salman. He had grown up in New Jersey with his mother and sister. As a teenager he had struck out for Pakistan to fight with Kashmiri separatists. He ended up in a training camp in Afghanistan, he said, run by a colonel from Pakistani intelligence.

  “They found out I was an American, and the ISI colonel flipped out!” Salman later told Charlie Santos, Miller’s business partner on the pipeline deal. Salman said he had been ordered to leave the training camp. He enlisted with the Taliban, who did not seem to mind having an American in their midst. “These guys are so pure, and they’re such good guys,” Salman said.

  He asked how the Knicks were doing. Santos felt sorry that he did not have much of a standings update.

  Miller had brought along a three-page, nonbinding agreement letter that he wanted the Taliban to sign. It would confirm the Taliban’s willingness to work with Unocal on the pipeline project. The leter outlined only a “preliminary basis for further discussions,” and it said that the pipelines could only go forward with “the establishment of a single, internationally recognized entity” running Afghanistan, a government “authorized to act on behalf of all Afghan parties.”21

  Miller and Santos explained that Unocal wanted to work with all Afghan factions. “But we want to dominate,” one of the Taliban’s negotiators replied.

  The Unocal group began to think that maybe the Taliban weren’t the village idiots everyone thought they were. They wanted the pipeline contract, but only on their terms and only if it could be had without any involvement of Ahmed Shah Massoud’s faction in Kabul, or any other Afghan rivals. Time, the Taliban’s negotiators seemed to believe, was on their side.

  Marty Miller gave up and drove west to meet with Taliban leaders in Herat. The long road from Kandahar was a potholed rut. Upon arrival the Taliban’s local governor welcomed Miller by looking him in the eye and asking menacingly, “Why don’t you convert to Islam?”

  On the long, grinding drive back, Taliban militia forced Miller’s convoy to stay overnight in a tiny mud hut along the highway. There was trouble on the road, and it was too dangerous to go farther in the dark. Other Afghan villagers had gathered at the checkpoint as well. They pressed around Miller, curious. Miller didn’t like the attention, so he climbed back into his truck, lay down on the seat, and strapped his Walkman to his ears, trying to escape into his music. After a few minutes he looked up and saw dozens of Afghan eyes pressed against the truck window, staring at him. He stayed inside his truck cab all night.

  The caravan stopped again briefly in Kandahar. The Taliban’s leaders still would not sign Unocal’s cooperation letter. Miller and his team climbed back in their pickups and left for Quetta.22 When they crossed into Pakistan, Miller climbed out of his truck, kissed the ground, and did a little dance of celebration. There were some places even a Texas wildcatter did not belong.

  18

  “We Couldn’t Indict Him”

  A CIA CASE OFFICER visited Marty Miller regularly at Unocal’s Sugarland, Texas, offices, usually after Miller had returned from a long overseas trip. Miller was not a CIA agent and did not take assignments, money, or instructions from the agency. But like some other American oil executives with access to the Middle East and Central Asia, he voluntarily provided briefings to the CIA’s Houston station. William Casey had revitalized the CIA’s contacts with American businessmen during the 1980s. He thought the agency overvalued its paid sources and missed out on the inside details that international businessmen picked up. Miller told the Houston officer about his negotiations in Turkmenistan and Pakistan, the gossip he overheard about corruption cases, and what he saw and heard when he traveled inside Afghanistan. The briefing sessions were dominated by Miller’s reports, but occasionally the CIA officer would provide some useful detail in exchange. At one stage the CIA became worried about threats to Unocal executives in Central Asia from Iranian intelligence operatives. The agency invited Miller to Langley for a briefing on how to manage his movements to reduce risk. Miller’s impression from his meetings was that the CIA was curious about Unocal’s Afghan pipeline plans but had no special interest in either the project or Afghanistan. In his efforts to win support for Unocal’s pipeline plan within the U.S. government, Miller maintained more active lobbying contacts at the White House and the State Department than at the CIA.1

  By early 1996 the agency was more estranged from its former Afghan and Pakistani contacts than at any time since the Soviet invasion in 1979. The U.S. ambassador in Islamabad, Tom Simons, was startled to find the CIA “had nothing” in Afghanistan. “They had taken out all their assets. They were basically past it.”2 Stinger missile recovery remained the only well-funded covert action program in the region. The Islamabad station did continue to collect intelligence on regional terrorism. Its officers tracked and mapped Afghan guerrilla training camps that supplied Islamist fighters in Kashmir. They continued to look for Mir Amal Kasi in the tribal territories along the Afghan border. But the liaison between the CIA’s Islamabad station and Pakistani intelligence—the spine of American covert action and intelligence collection in the region for fifteen years—had cracked. Javed Qazi had been replaced as ISI chief by another mainstream general, Naseem Rana, a Punjabi officer with a background in the signals corps. Some of the Americans who dealt with him found Rana a dull-minded time server who was unwilling to go out of his way to help the United States. Pakistani intelligence offered little cooperation in the search for Karachi terrorists who murdered two Americans in 1995. After a raid on the Kasi family home in Quetta turned up nothing because of faulty intelligence supplied by the Americans, ISI essentially shut down its operations on that case. If the CIA developed hard, convincing evidence about Kasi’s location—evidence that Pakistan could confirm—then ISI would assist in his capture, Rana said. But that was about it. Commission payments to ISI for recovered Stingers provided a thin basis for cooperation, but meetings between the CIA and Pakistani intelligence in Rawalpindi were infrequent and desultory compared to the past.3

  Gary Schroen, the longtime CIA Afghan hand who had served two previous tours in Islamabad, arrived as station chief in January 1996. He told colleagues that the Unocal pipeline project was a fool’s errand and that he was not going to pay any attention to it. The pipeline would never be built, Schroen predicted. Besides, the Islamabad station no longer had Afghanistan on its Operating Directive. This bureaucratic designation meant that Schroen and his case officers had no authority to collect intelligence on the Taliban’s strengths, sources of supply, or military prospects. Nor could they develop similar intelligence about Hekmatyar’s militia or Massoud’s Kabul government. The Islamabad station could recruit Afghan agents if they were reporting on terrorism, drugs, or Stinger missiles. But the default assignment of the Afghan account to Langley created occasional confusion within the CIA about how to track the spillover effects of Afghanistan’s civil war.4

 
CIA headquarters was distracted by scandal, shrinking budgets, a wave of early retirements, controversies in Congress, and leadership turmoil in the director’s office. Not since the late 1970s had so many career agency officers felt so miserable about the place.

  Clinton fired James Woolsey in early 1995, after the Aldrich Ames spy case broke. Ames had worked for Russia inside Langley headquarters for years, and his betrayal had gone undetected. The president struggled to find a successor and finally turned to John Deutch, then deputy secretary of defense, who told Clinton adamantly that he did not want the CIA job. Clinton insisted; there was no one else available who could win confirmation, he said. An MIT-educated chemist who had first come to Washington during the 1960s as a “whiz kid” analyst in Robert McNamara’s Pentagon, Deutch was a large, bearish man with an ample belly. He had the independent, inquiring, self-certain mind of an accomplished scientist. He could be warm, sloppy, and professorial but also caustic, dismissive, and arrogant. He was happy at the Pentagon, where he worked with a friend and mathematician, William Perry. He had watched James Woolsey, whom he regarded as a very able man, fail spectacularly at Langley, and he had no desire to follow him. Yet once persuaded by the president, Deutch decided to hit the CIA with all of the force he could muster. Congress and the press were outraged over the Ames case. Senator Daniel Patrick Moynihan, a longtime CIA skeptic, had introduced legislation to abolish the agency and fold its role into other departments. Even the CIA’s supporters could not understand how the clues about Ames’s treachery—his outlandish personal spending, for instance—had been missed. Deutch joined the reformers: He pledged at his confirmation hearing to change the CIA “all the way down to the bare bones.”5

  Deutch openly described himself as “a technical guy, a satellite guy, a SIGINT guy,” referring to “signals intelligence,” or the art of communications intercepts. He used his early budget requests at Langley to direct more money proportionately to other agencies in the intelligence community, such as the National Reconnaissance Office at the Pentagon and the National Security Agency. He thought the CIA’s historical strength was scientific and technical intelligence collection, and he wanted to concentrate on that. He was not impressed with the agency’s human spying operations. He believed that the leadership of the Directorate of Operations had to be reformed. His sense was that the CIA’s spies were just not very good anymore at their core job of agent recruitment and intelligence collection. They had forgotten the basics of espionage. They were not living up to their own professional standards, and he was not afraid to tell them so. “From what I know, the junior officers are waiting for some new direction,” Deutch said publicly. “Now, I may be unhappily surprised.”6

  He was. Many of the CIA’s career officers revolted against Deutch’s change message. They saw his management reform campaign as just the latest wave in a series of attacks against the agency’s core mission and culture. To them President Clinton seemed indifferent about the CIA’s health. The agency’s budget continued to shrink. In mid-1995 there were only a dozen new case officers being trained at the Farm as career spies. The Directorate of Operations now had fewer than eight hundred case officers worldwide, about a 25 percent decline from the peak years of the Cold War. Stations had closed not only in Afghanistan but across the Third World. There was a strong sense in the Directorate of Operations that the CIA was getting rolled in the budget process by the Pentagon and the FBI. After the Ames case, internal investigations into other possible spies operating at Langley placed dozens of case officers under suspicion, contributing to an atmosphere of distrust and uncertainty. When Deutch’s new managers arrived, they emphasized gender and racial diversity as a prime CIA hiring goal, a mission that angered and dismayed the many white males among the agency’s veterans. New management techniques promoted open criticism of supervisors, discussions about the CIA’s purpose, focus groups, more interaction with the media—“California hot tub stuff,” as one unhappy veteran called it. To achieve personnel reductions without firing anyone, CIA managers had to look for experienced officers who were vested enough in their pensions to be able to retire early without hardship. They sought out such veterans and encouraged them to leave. The retirements became wrenching and disruptive.7

  On the day he accepted early departure, longtime Soviet analyst Fritz Ermarth filled out paperwork with his retirement counselor, an old acquaintance he had known since the days of CIA directors Stansfield Turner and William Casey. Ermarth posed the kind of question that he used to ask about the Soviet bureaucracies he analyzed: “Look, you process four hundred to five hundred people a year through this little cubicle, right? What’s your portrait of the place?”

  The counselor’s eyes filled with tears. “I’ve never seen it so bad,” she said, as Ermarth recalled it. He asked what she meant.

  “Everybody says it’s hard to put your finger on it,” she replied, “but it’s the growth in the importance of stuff that shouldn’t matter relative to stuff that should.”8

  THE CIA’S COUNTERTERRORIST CENTERE began to emerge as a modest exception to the agency’s downward trend. For the first two years of the Clinton presidency, budgeting and policy making about terrorism had been dispersed and confused. The shock of the Oklahoma City bombing in the spring of 1995 created a new sense of urgency at the National Security Council, however. The bombers turned out to be a domestic cell of antigovernment militia. But their audacious strike coincided with a shocking chemical weapons attack by a Japanese-based cult in Tokyo.White House terrorism analysts believed the Japanese case showed that the United States was vulnerable to terrorists using weapons of mass destruction. Spurred by Clinton, the National Security Council organized its first terrorism policy review during the early months of 1995.

  In June, Clinton signed Presidential Decision Directive-39, classified Secret, titled “U.S. Policy on Counterterrorism.” The document echoed the presidential directive that President Reagan had signed during the last great wave of anti-American terrorism during the mid-1980s. It was also the first official recognition by any American president of the danger posed to the United States by terrorists who acquired nuclear, chemical, or biological weapons.9

  The CIA was instructed to undertake “an aggressive program of foreign intelligence collection, analysis, counterintelligence, and covert action.” If necessary, CIA operations would seek to return terrorist suspects “by force … without the cooperation of the host government” so that the accused could face justice in American courts.

  “The acquisition of weapons of mass destruction by a terrorist group, through theft or manufacture, is unacceptable,” the directive continued. “There is no higher priority than preventing the acquisition of this capability or removing this capability from terrorist groups potentially opposed to the U.S.”10

  On paper, at least, American policy was now more forceful and clearly stated than it had been in years. The document also centralized authority on counterterrorism policy at the White House for the first time. The challenge now was to put the words into practice.

  ——————

  IN JANUARY 1996 the CIA’s Counterterrorist Center opened a new office to track Osama bin Laden. The agency had never before dedicated a unit of this kind to a single terrorist. Formally known as the “bin Laden Issue Station” and code-named “Alec,” the group leased space in a suburban Virginia office park just a few miles from CIA headquarters. Employing about twelve staff members, it was designated a “virtual station.” This meant that within the CIA’s budgeting and cable routing systems, the unit would have the administrative status, privileges, and autonomy enjoyed by more traditional stations abroad. The idea was born from discussions in the Counterterrorist Center’s senior management group. Bin Laden was still seen by CIA analysts primarily as a money man, but he was an emerging symbol of the new mobility of international terrorism. National Security Adviser Tony Lake, who approved the bin Laden unit at the CIA, recalled that he realized the Saudi had become an important t
errorist when classified memos started referring to him by the acronym “UBL” (which referred to a spelling of bin Laden’s transliterated first name as Usama). In Washington having an acronym was the ultimate sign of importance, Lake recalled sardonically. Because he operated across borders, bin Laden presented challenges to the CIA’s old system of country-based intelligence collection. The CIA’s managers wanted to experiment with a new kind of unit, a prototype that might be used against other transnational targets. They would fuse intelligence disciplines into one office—operations, analysis, signals intercepts, overhead photography, and so on. The National Security Agency had tapped into bin Laden’s satellite telephone and kept track of his international conversations. These intercepts could be used by the new station to track his payments and connections in multiple countries.11

  They chose bin Laden because by early 1996 there was a rising recognition of his importance, both at the CIA’s Counterterrorist Center and at the White House. The unit’s first project was to develop a strategic picture of bin Laden’s activity. Some of the new focus on bin Laden came from Richard Clarke, a forceful career civil servant who in the summer of 1995 had been appointed Clinton’s counterterrorism director, working from the National Security Council under the authorities spelled out in PDD-39. In addition, classified evidence about bin Laden was piling up, circulating in cables throughout the intelligence community. The reporting from the CIA’s Khartoum station was by now voluminous. Bin Laden’s name surfaced continually in reports from Egypt, Algeria, Tunisia, Israel, and elsewhere. As one regular reader of these cables recalled, it seemed as if every other cable about terrorism from North Africa contained the phrase “Osama bin Laden, financier of terrorists.” The CIA now viewed bin Laden as “one of the most significant financial sponsors of Islamic extremist activities in the world,” as a rare public statement put it. There was some new money available for the CIA’s counterterrorism budget by fiscal 1996. Tony Lake chaired an interagency meeting that approved spending it on the CIA’s virtual bin Laden station. Richard Clarke said later that he asked the CIA and the Pentagon to develop plans for “operating against” al Qaeda in Sudan, instead of merely collecting intelligence, but that neither department “was able successfully to develop a plan.” Operators inside the virtual station began drafting plans to capture bin Laden early on, but none of these ideas was approved or carried forward by superiors or the White House. The agency’s plan offered a way to try something new: “Let’s yank on this bin Laden chain and see what happens,” as one participant recalled.12