The CIA Develops a Capture Plan

Initially, the DCI's Counterterrorist Center and its Bin Ladin unit considered a plan to ambush Bin Ladin when he traveled between Kandahar, the Taliban capital where he sometimes stayed the night, and his primary residence at the time, Tarnak Farms. After the Afghan tribals reported that they had tried such an ambush and failed, the Center gave up on it, despite suspicions that the tribals' story might be fiction. Thereafter, the capture plan focused on a nighttime raid on Tarnak Farms.¹⁷

A compound of about 80 concrete or mud-brick buildings surrounded by a 10-foot wall, Tarnak Farms was located in an isolated desert area on the outskirts of the Kandahar airport. CIA officers were able to map the entire site, identifying the houses that belonged to Bin Ladin's wives and the one where Bin Ladin himself was most likely to sleep. Working with the tribals, they drew up plans for the raid. They ran two complete rehearsals in the United States during the fall of 1997.¹⁸

By early 1998, planners at the Counterterrorist Center were ready to come back to the White House to seek formal approval. Tenet apparently walked National Security Advisor Sandy Berger through the basic plan on February 13. One group of tribals would subdue the guards, enter Tarnak Farms stealthily, grab Bin Ladin, take him to a desert site outside Kandahar, and turn him over to a second group. This second group of tribals would take him to a desert landing zone already tested in the 1997 Kansi capture. From there, a CIA plane would take him to New York, an Arab capital, or wherever he was to be arraigned. Briefing papers prepared by the Counterterrorist Center acknowledged that hitches might develop. People might be killed, and Bin Ladin's supporters might retaliate, perhaps taking U.S. citizens in Kandahar hostage. But the briefing papers also noted that there was risk in not acting. "Sooner or later," they said, "Bin Ladin will attack U.S. interests, perhaps using WMD [weapons of mass destruction]."¹⁹

Clarke's Counterterrorism Security Group reviewed the capture plan for Berger. Noting that the plan was in a "very early stage of development," the NSC staff then told the CIA planners to go ahead and, among other things, start drafting any legal documents that might be required to authorize the covert action. The CSG apparently stressed that the raid should target Bin Ladin himself, not the whole compound.²⁰

The CIA planners conducted their third complete rehearsal in March, and they again briefed the CSG. Clarke wrote Berger on March 7 that he saw the operation as "somewhat embryonic" and the CIA as "months away from doing anything."²¹

"Mike" thought the capture plan was "the perfect operation." It required minimum infrastructure. The plan had now been modified so that the tribals would keep Bin Ladin in a hiding place for up to a month before turning him over to the United States-thereby increasing the chances of keeping the U.S. hand out of sight. "Mike" trusted the information from the Afghan network; it had been corroborated by other means, he told us. The lead CIA officer in the field, Gary Schroen, also had confidence in the tribals. In a May 6 cable to CIA headquarters, he pronounced their planning "almost as professional and detailed . . . as would be done by any U.S. military special operations element." He and the other officers who had worked through the plan with the tribals judged it "about as good as it can be." (By that, Schroen explained, he meant that the chance of capturing or killing Bin Ladin was about 40 percent.) Although the tribals thought they could pull off the raid, if the operation were approved by headquarters and the policymakers, Schroen wrote there was going to be a point when "we step back and keep our fingers crossed that the [tribals] prove as good (and as lucky) as they think they will be."²²

Military officers reviewed the capture plan and, according to "Mike," "found no showstoppers." The commander of Delta Force felt "uncomfortable" with having the tribals hold Bin Ladin captive for so long, and the commander of Joint Special Operations Forces, Lieutenant General Michael Canavan, was worried about the safety of the tribals inside Tarnak Farms. General Canavan said he had actually thought the operation too complicated for the CIA-"out of their league"-and an effort to get results "on the cheap." But a senior Joint Staff officer described the plan as "generally, not too much different than we might have come up with ourselves." No one in the Pentagon, so far as we know, advised the CIA or the White House not to proceed.²³

In Washington, Berger expressed doubt about the dependability of the tribals. In his meeting with Tenet, Berger focused most, however, on the question of what was to be done with Bin Ladin if he were actually captured. He worried that the hard evidence against Bin Ladin was still skimpy and that there was a danger of snatching him and bringing him to the United States only to see him acquitted.²⁴

On May 18, CIA's managers reviewed a draft Memorandum of Notification (MON), a legal document authorizing the capture operation. A 1986 presidential finding had authorized worldwide covert action against terrorism and probably provided adequate authority. But mindful of the old "rogue elephant" charge, senior CIA managers may have wanted something on paper to show that they were not acting on their own.

Discussion of this memorandum brought to the surface an unease about paramilitary covert action that had become ingrained, at least among some CIA senior managers. James Pavitt, the assistant head of the Directorate of Operations, expressed concern that people might get killed; it appears he thought the operation had at least a slight flavor of a plan for an assassination. Moreover, he calculated that it would cost several million dollars. He was not prepared to take that money "out of hide," and he did not want to go to all the necessary congressional committees to get special money. Despite Pavitt's misgivings, the CIA leadership cleared the draft memorandum and sent it on to the National Security Council.²⁵

Counterterrorist Center officers briefed Attorney General Janet Reno and FBI Director Louis Freeh, telling them that the operation had about a 30 percent chance of success. The Center's chief, "Jeff," joined John O'Neill, the head of the FBI's New York Field Office, in briefing Mary Jo White, the U.S. Attorney for the Southern District of New York, and her staff. Though "Jeff" also used the 30 percent success figure, he warned that someone would surely be killed in the operation. White's impression from the New York briefing was that the chances of capturing Bin Ladin alive were nil.²⁶

From May 20 to 24, the CIA ran a final, graded rehearsal of the operation, spread over three time zones, even bringing in personnel from the region. The FBI also participated. The rehearsal went well. The Counterterrorist Center planned to brief cabinet-level principals and their deputies the following week, giving June 23 as the date for the raid, with Bin Ladin to be brought out of Afghanistan no later than July 23.²⁷

On May 20, Director Tenet discussed the high risk of the operation with Berger and his deputies, warning that people might be killed, including Bin Ladin. Success was to be defined as the exfiltration of Bin Ladin out of Afghanistan.²⁸ A meeting of principals was scheduled for May 29 to decide whether the operation should go ahead.

The principals did not meet. On May 29, "Jeff" informed "Mike" that he had just met with Tenet, Pavitt, and the chief of the Directorate's Near Eastern Division. The decision was made not to go ahead with the operation. "Mike" cabled the field that he had been directed to "stand down on the operation for the time being." He had been told, he wrote, that cabinet-level officials thought the risk of civilian casualties-"collateral damage"-was too high. They were concerned about the tribals' safety, and had worried that "the purpose and nature of the operation would be subject to unavoidable misinterpretation and misrepresentation-and probably recriminations-in the event that Bin Ladin, despite our best intentions and efforts, did not survive."²⁹

Impressions vary as to who actually decided not to proceed with the operation. Clarke told us that the CSG saw the plan as flawed. He was said to have described it to a colleague on the NSC staff as "half-assed" and predicted that the principals would not approve it. "Jeff " thought the decision had been made at the cabinet level. Pavitt thought that it was Berger's doing, though perhaps on Tenet's advice. Tenet told us that given the recommendation of his chief operations officers, he alone had decided to "turn off" the operation. He had simply informed Berger, who had not pushed back. Berger's recollection was similar. He said the plan was never presented to the White House for a decision.³⁰

The CIA's senior management clearly did not think the plan would work. Tenet's deputy director of operations wrote to Berger a few weeks later that the CIA assessed the tribals' ability to capture Bin Ladin and deliver him to U.S. officials as low. But working-level CIA officers were disappointed. Before it was canceled, Schroen described it as the "best plan we are going to come up with to capture [Bin Ladin] while he is in Afghanistan and bring him to justice."³¹ No capture plan before 9/11 ever again attained the same level of detail and preparation. The tribals' reported readiness to act diminished. And Bin Ladin's security precautions and defenses became more elaborate and formidable.

At this time, 9/11 was more than three years away. It was the duty of Tenet and the CIA leadership to balance the risks of inaction against jeopardizing the lives of their operatives and agents. And they had reason to worry about failure: millions of dollars down the drain; a shoot-out that could be seen as an assassination; and, if there were repercussions in Pakistan, perhaps a coup. The decisions of the U.S. government in May 1998 were made, as Berger has put it, from the vantage point of the driver looking through a muddy windshield moving forward, not through a clean rearview mirror.³²