Ghost Wars (64 page)

The CIA did periodically obtain evidence that terrorist groups were interested in weapons of mass destruction. Tenet did not talk about it in public, but bin Laden now figured in this alarming, if fragmentary, CIA reporting. Late in 1996 a former bin Laden aide and courier, Jamal al-Fadl, entered an American witness protection program and provided detailed accounts of bin Laden’s earlier operations in Sudan. The CIA was involved in al-Fadl’s secret debriefings. Al-Fadl said that bin Laden had authorized attempts to buy uranium that might be used to fashion a nuclear bomb. This effort had failed as far as al-Fadl knew, but if he was telling the truth—and al-Fadl passed the polygraph tests he was given—his testimony suggested the scale of bin Laden’s ambitions. The CIA also had reports of contacts between bin Laden and Iraqi intelligence agents dating back to bin Laden’s years in Sudan, and there were some fragmentary indications that these Iraqi contacts had involved training in the development and use of chemical weapons.
33
Still, neither the White House nor the CIA as yet had any covert action program targeting bin Laden that went beyond intelligence collection and analysis. The CIA’s Counterterrorist Center was trying to watch bin Laden. Its leaders had not yet seriously attempted to arrest or kill him.

That planning was about to begin.

PART THREE

THE DISTANT ENEMY

January 1998 to September 10, 2001

21

“You Are to

Capture Him Alive”

THE FIRST FORMAL CIA PLAN to capture or kill Osama bin Laden began as a blueprint to arrest Mir Amal Kasi, the Baluchi migrant who had shot up the entrance to the agency’s headquarters in 1993.

Kasi remained a fugitive in the Afghanistan-Pakistan-Iran borderlands. The CIA’s Counterterrorist Center at Langley asked the Islamabad station for help recruiting agents who might be able to track him down. The station identified and contacted a family-based group of Afghan tribal fighters whose leadership had formal military training and who had worked for the CIA during the anti-Soviet jihad. Case officers met with the group and won their agreement to come back on the agency payroll to hunt for Kasi. At Langley, officers in the Counterterrorist Center’s Kasi cell secured budget approval for the recruitment. The headquarters unit shipped hundreds of thousands of dollars in cash, AK-47 assault rifles, land mines, motorcycles, trucks, secure communications equipment, and electronic listening devices to put its new Afghan agents into business. Langley also supplied mobile beacons that could be used to pinpoint the exact location of buildings by connecting to satellites hovering miles overhead. The technology would allow an American counterterrorism team to swarm an obscure location quickly once it was lit up by the Afghan agents. The tribal team had been code-named GE/SENIOR during the anti-Soviet years. Now they were dubbed by a new cryptonym, FD/ TRODPINT. The suddenly enriched and provisioned Afghans set up residences around Kandahar, traveled back and forth to Pakistan, and began to track leads that might eventually take them to Kasi. In effect they had signed up as lethal, exceptionally well paid CIA bounty hunters.
1

There were clear authorities for the recruitment under U.S. law. Kasi had been indicted for murder in the United States. Under federal law such fugitives could be arrested abroad and returned to the United States for trial. By collecting intelligence overseas about a suspect’s whereabouts, the CIA could aid such an arrest under standing legal authorities approved by the president. Under these federal rules, the role played by CIA case officers and paid Afghan agents in tracking Kasi down need never be known. If the tracking team found Kasi in Pakistan, they were to contact the CIA station in Islamabad. Case officers would then attempt to work with Pakistani intelligence and police to make an arrest without revealing the existence of their paid Afghan agents.

A trickier scenario would arise if the tribal agents found Kasi hiding in southern Afghanistan, however. The Taliban controlled most of the traditional Baluch territory where Kasi was presumed to be moving. Given the record of stilted, sometimes bizarre contacts between American officials and the Taliban’s Kandahar leadership, it was impossible to conceive of a cooperative approach with them. Legally, the United States did not even recognize the Taliban. Yet the Rabbani-Massoud government, which did have tentative legal standing, had no practical authority in Taliban country. If the CIA was going to take Kasi into custody in that area, it was going to have to find a way to do so on its own.

Agency case officers in Islamabad met with their tribal team to develop a formal, specific plan to capture Kasi in southern Afghanistan and fly him to the United States for trial. The plan would require the Afghan agents to hold Kasi securely in place long enough for an American arrest team to fly in secretly, bundle the fugitive aboard an airplane or helicopter, and lift off safely for the United States.

Because of their military training, the tribal agents talked convincingly about their ability to mount such a capture operation. The Afghan team worked well with maps. They had a sense of time and military sequence. They could identify assembly points, rally points, escape routes. One question was how to insert an American squad into Afghanistan if the tracking team located and detained Kasi on its own. The CIA’s case officers provided their Afghan recruits with specifications for a suitable landing strip that could be prepared in advance. The chosen desert ground had to be hard and stable enough to support an aircraft landing and takeoff. It had to be secure from Taliban forces, preferably in a lightly populated and isolated valley. It had to be adequate for pilot navigation. The Afghan agents struck out on their motorcycles around Kandahar. They carried satellite measuring devices to pinpoint coordinates for possible airstrip sites. When they found a candidate location, they transmitted the data to Islamabad, and the station then ordered satellite photography to examine the site’s parameters from above. Eventually the CIA found a remote strip that looked suitable, at least from the vantage of satellites.

The CIA and the Pentagon did not typically send American officers into harm’s way based solely on satellite pictures and the investigations of paid Afghan recruits.What if the dirt at the landing site proved too soft despite the agents’ assurances, and the American team’s plane got stuck in the sand?

At Langley the Counterterrorist Center proposed and won approval for what CIA officers call a “black op,” a secret operation classified at the highest possible level. The mission would both confirm the desert landing site’s suitability and rehearse for the day when Kasi was actually in agent custody. A special operations team flew secretly into Afghanistan. Without Pakistan’s knowledge, they mounted a nighttime low-level flight, tested the chosen landing zone marked by the tribal agents, found it satisfactory, double-checked its satellite coordinates, and withdrew. The CIA’s Afghan capture plan for Mir Amal Kasi was now as ready as it could be for launch.

But month after month passed during 1996 and early 1997, and Kasi could not be found. The CIA’s deteriorating relationship with Pakistani intelligence was one factor; the agency received little access to Pakistani police resources in the borderlands. The sprawling, centuries-rooted web of clan and tribal protection available to any Baluch in trouble in the territory of his birth was perhaps a greater problem. The CIA’s case officers sought to combat Kasi’s call on clan loyalty with appeals to greed. They offered multimillion-dollar rewards both openly and privately to anyone willing to reveal Kasi’s whereabouts. But for months there were no takers. Under traditional Baluch revenge codes, anyone exposed as Kasi’s betrayer risked not only his own life but his family’s as well. For a while the CIA picked up rumors that Kasi was staying in a massive fortress compound near the Afghan border, but the agency could not persuade Pakistani police to move against the place. The operation would have been unusually difficult because the compound was heavily defended. CIA officers tried a technical solution: They rigged a special television with a roving camera that looked out from behind the TV screen. They arranged to deliver the set inside the compound, hoping to catch a picture of Kasi on film. The operation turned up nothing, however. It was never clear whether Kasi had ever been inside the place.

Finally their luck turned. In late May 1997 a Baluch man walked into the U.S. consulate in Karachi and told a clerk he had information about Kasi. He was taken to a young female CIA officer who was chief of base in Karachi (an agency “base” is a subunit of a countrywide station). She interviewed the informant and concluded he was credible. The CIA officer and the FBI’s attaché in Pakistan, Scott Jessie, arranged more interviews. The source claimed that about two years earlier Kasi had been placed under the protection of a prominent Baluch tribal leader; the pair had become confidants and business partners, and traveled together frequently. Now, the source explained, the tribal leader had decided to sell out Kasi to the U.S. government in exchange for the reward money. The source handed over an application for a Pakistani driver’s license filled out by Kasi under an alias; it contained a photo and a thumbprint that confirmed they had their man. The tribal leader who had befriended Kasi flew to Karachi and worked out an arrest plan with the CIA and the FBI. The tribal chief would be visiting a central Pakistani town called Dera Ghazi Khan on business in mid-June. He promised to lure Kasi to the Shalimar Hotel where the FBI could arrest him.

Naseem Rana, the director of Pakistani intelligence, had repeatedly told CIA station chief Gary Schroen that if he could locate Kasi, Pakistani police would help arrest him. Now Schroen and Jessie met with ISI officers and laid out their specific plan. They asked the Pakistanis to fly teams of CIA officers and FBI agents on a military plane to Multan, the largest Pakistani city near Dera Ghazi Khan. ISI would then provide ground transportation to the Shalimar and secure the perimeter while the FBI went in. Then they would all fly back to Islamabad where ISI would allow Kasi to be flown immediately to the United States. Rana agreed to the plan in its entirety.

The CIA’s Karachi base chief and the tribal leader flew into Multan for the big day. Just before dawn on June 15, 1997, Kasi’s betrayer knocked on the hotel room door and shouted that it was time to get up for dawn prayers. FBI agents stood at his shoulder. Schroen and two CIA colleagues waited outside, holding a secure satellite radio linked to Langley headquarters. FBI Special Agent Brad Garrett kicked through the door, straddled Kasi on the floor, pressed the suspect’s left thumb onto an ink pad, studied the result with a magnifying glass, and declared exultantly, “It’s a match!” They raced to the Multan airport in six sport utility vehicles, with gunmen from Pakistani intelligence hanging from the windows. On the tarmac next to a CIA helicopter an agency officer connected Schroen’s secure radio to Langley where Tenet and other senior officials had gathered to monitor the operation. “This is Red Light Zulu,” Schroen announced, declaring his call sign. “The package was successfully picked up and is safely bundled and being loaded onto an aircraft for movement to Islamabad. All personnel on the team are safe. This was a totally successful mission.”
2

A case that ranked first at CIA headquarters had finally been closed. George Tenet summoned five hundred employees to the Langley auditorium and arranged a closed-circuit television broadcast throughout headquarters. He played a recording of Schroen’s “Red Light Zulu” message for the entire CIA staff. “No terrorist should sleep soundly as long as this agency exists,” Tenet announced triumphantly. He urged his colleagues to give one another high-fives and hugs, and to “have a cocktail before noon.”
3

In the heady weeks that followed a question arose inside the CIA’s Counterterrorist Center about what would now become of their elaborately equipped and financed TRODPINT tracking team assets. It seemed a shame to just cut them loose. A few flimsy U.S. government partitions away from the Kasi tracking team stood the small cluster of analysts and operators who made up the bin Laden issue unit. (After a relatively brief life in a Virginia office park, the station had been reincorporated into the headquarters of the Counterterrorist Center.) By the summer of 1997 the unit was reporting regularly to policy makers in classified channels about threats issued by bin Laden against American targets, especially American military forces stationed in Saudi Arabia. The CIA continued to describe bin Laden as an active, dangerous financier of Islamist extremism in Egypt, Sudan, Algeria, and Kashmir.

Yet the CIA had few ways to keep track of bin Laden on its own. Now the tribal team beckoned. The paid, well-organized Afghan agents could monitor or harass bin Laden up close, under direct CIA control.

Paul Pillar, the Princeton-educated analyst who had helped shape the CIA’s thinking about the terrorist threat in the Middle East during the early 1990s, was now the center’s deputy director. His superior, the Counterterrorist Center’s director in the summer of 1997, Jeff O’Connell, was a veteran from the Directorate of Operations who had experience in Yemen, knew Egypt well, and had long studied the threat of Islamist extremism then rising in the Arab world.
4
He approved a plan that summer to transfer the Afghan agent team from the Kasi cell to the bin Laden unit, which had been developing draft plans to attack bin Laden’s facilities and financial assets since 1996.

The agents’ new CIA controllers modified the Kasi capture plan so it could be used to seize bin Laden and bring him to justice. At the CIA’s Islamabad station this initiative arrived that summer of 1997 in the form of cables from Langley authorizing meetings with the tribal team leadership to explain that if they wanted to remain on the agency payroll, they now had to go after bin Laden. The Afghan TRODPINT team agreed.

As a target for capture, bin Laden was an easier mark than Kasi. At least they knew for certain where bin Laden lived some of the time: in the compounds provided by Mullah Omar in and around Kandahar. As Unocal’s executives and liaisons had discovered early in 1997, bin Laden moved freely through the Taliban capital. His bodyguard and some of his wives and children lived openly near the Kandahar airport.