PAUL PILLAR ARRIVED AS CHIEF of analysis at the CIA’s Counterterrorist Center six weeks after the World Trade Center bombing. He was a tall, lanky man with a nervous blink and the careful, articulate voice of a university professor. A U.S. army officer in Vietnam, he had degrees from Dartmouth and Oxford, and a doctorate from Princeton. Fast-tracked in management and intelligence analysis after he joined the CIA, he served as executive assistant to CIA director William Webster, a position often set aside at Langley for rising stars. Pillar reflected the high-minded traditions of the CIA’s analytical wing, the Directorate of Intelligence. He was not an Arabist, but he had studied political Islam and the Middle East. He was a manager and an intellectual, an author of books and academic journal articles. From within the Counterterrorist Center he would emerge during the next six years as one of the CIA’s most influential terrorism analysts.1
Initially, Pillar was as stumped as the FBI was by the World Trade Center case. The first group of suspects arrested in the New York area were a diverse, bumbling crew. It was easier to imagine them as pawns of some hidden foreign government plot than as an independent terrorist cell. Gradually, as the FBI’s evidence accumulated, a new theory of the case began to emerge. Informants quickly identified the blind Egyptian preacher Sheikh Rahman as a source of inspiration for the World Trade Center attack and several thwarted bombings of New York landmarks. The CIA’s analysts began to look more closely at cross-border Islamist radicalism emanating from Egypt and its neighbors.
An Islamic political revival had swept through the Arab countries of North Africa during the previous four years. A violent offshoot of the Muslim Brotherhood in Egypt, the Islamic Group, had opened an assassination and bombing campaign against the secular government of Hosni Mubarak. The Islamic Group’s cadres hailed from the impoverished, long-radicalized Upper Nile region, and their campaign revived a decades-old tradition of Islamist violence in Egypt. But the group also seemed to be newly stimulated by returning veterans from the Afghan jihad. The same was true in Algeria. There the Muslim Brotherhood–linked Islamic Salvation Front captured the political imagination of Algeria’s poor and, increasingly, its angry middle classes, who saw their secular, socialist leaders as corrupt and politically exhausted. After the government aborted a 1991 election because it appeared the Islamists would win, young militants, some of them veterans of the Afghan jihad, went underground and formed a new violent resistance called the Armed Islamic Group. They opened a terror campaign against the government. Hundreds of Algerian civilians died each month in bombings, massacres, and assassinations carried out by both sides.2
Pillar and other CIA analysts, along with station chiefs in Cairo, Algiers, and Tunis, studied and debated these insurgencies intently in the months after the World Trade Center attack. They asked: What was the connection between these violent, national Islamist groups and terrorists who might threaten the United States or its allies? What policy should the United States adopt toward the Egyptian and Algerian Islamists? Should it regard all Islamic fundamentalists as dangerous, or should Washington reach out to the peaceful wings of the Muslim Brotherhood, while attempting to isolate and repress its violent offshoots? Should the United States encourage democratic elections even in countries such as Algeria or Egypt where the Islamists might win? How could Washington be sure that Islamists would continue with a democratic system after they won power?
Pillar and his colleagues saw the fall of the Soviet Union in 1991 and the fall of the Shah of Iran in 1979 as models of political failure from which they hoped to learn. In both of those historical cases corrupt, failing governments with little credibility had faced popular rebellions, tried to reform themselves, and collapsed anyway. Pillar thought the lesson might be that you had to avoid half measures: A government under violent siege should either strike back ruthlessly or open up its political system completely. Still, he felt the Algerians had made a terrible mistake by canceling their election and driving the Islamists underground. They had strengthened the extremists and isolated the Muslim Brotherhood’s peaceful politicians.
Senior intelligence analysts and policy makers at the State Department and the National Security Council were also torn. Algeria and Tunisia, although not close American allies, were secular bulwarks, increasingly pro-Western on security issues. Egypt, the most populous and historically influential country in the Arab world, was one of America’s closest allies, the second largest recipient of American aid after Israel, and a crucial partner with Washington in the Israeli-Palestinian peace process. Pillar and other CIA analysts believed the United States should do everything possible to shore up Mubarak’s government against the Islamists despite Mubarak’s obvious failings.
Yet the agency’s analysts remembered the experience of the Iranian revolution, where the CIA and the White House had clung to a failing despotic ally for too long and, by doing so, had deprived themselves of a chance to work constructively with Iran’s new revolutionary Islamic government. Pillar saw the Muslim Brotherhood in Jordan as a peaceful Islamist movement that was ready to participate in mainstream politics even though it voiced a radical philosophy. Perhaps the Muslim Brotherhood in other countries could be coaxed toward peaceful democracy.
Participants in these intelligence and policy debates during the first Clinton term recall them as fractured, disorganized, and inconclusive. Tony Lake had announced that the expansion of democracy worldwide would be a preeminent American goal during the 1990s. But with Islamist violence now raging in Algeria and Egypt, neither Lake nor Clinton was prepared to make a priority of urging democratic elections in the Arab world. The American embassy in Cairo reached out cautiously to the less violent leaders of the Muslim Brotherhood, but the dialogue never went very far.3
The most detailed intelligence collected by the CIA about radical Islamic movements in the Middle East during this early period came from its stations in Egypt, Algeria, Tunisia, and Israel. The CIA maintained a daily liaison with Egyptian intelligence and internal security forces. The agency’s Tunis station developed a similar liaison with Tunisian security forces as they cracked down against a Muslim Brotherhood–inspired Islamist movement. The CIA had sent its first declared station chief to Algiers in 1985 and maintained a working relationship with Algerian security forces even as they plunged into a bloody civil war. In all three countries the station chiefs recorded and cabled to Langley detailed alarmist accounts from Arab intelligence and police chiefs about the rising danger of Islamic radicalism. The North African officers complained repeatedly about the role of returning veterans of the Afghan jihad, the flow of Saudi Arabian funds, and the sanctuary available to violent radicals along the Pakistan-Afghanistan frontier. They complained also about the willingness of Britain, France, Germany, Sweden, and Denmark to grant asylum to exiled Islamist leaders.4
There was a clear pattern of international cooperation among the Islamic radicals tracked by the CIA’s North African stations. Tunisian security forces captured clandestine weapons moving by camel caravan from Sudan across the Sahara desert to Algeria. For the Tunisians during the months after the World Trade Center bombing “there was no other issue” to discuss with the CIA other than the threat of border-hopping Islamic radicals, recalled Whitley Bruner, then the Tunis station chief.5
Yasser Arafat and the leadership of the Palestinian Liberation Organization grew equally alarmed by the rise of these Muslim Brotherhood–inspired networks. As it embraced peace negotiations with Israel, the PLO faced a rising challenge from Hamas, the Palestinian branch of the Muslim Brotherhood. The PLO collected intelligence about Hamas’s fund-raising in Saudi Arabia, its religious schools in Yemen, and its gunrunning networks in Sudan. Terrorists with a violent Hamas offshoot, called the Palestinian Islamic Jihad, had clustered around an exiled Saudi financier named Osama bin Laden, the PLO informed the CIA’s station in Tel Aviv. The PLO hoped the CIA would join them in battle against the Islamists, disrupting Hamas.6
This early CIA reporting was tarnished by the poor reputations of its sources. Mubarak in Egypt, the massacre-sponsoring secret police in Algeria, the police state technocrats in Tunisia, and the corrupt leaders of the PLO all had self-serving reasons to exaggerate the dangers of their Islamic radical opponents. North Africa’s secular Arab governments were undemocratic and unpopular. The Islamists, some of them peaceful, had challenged their legitimacy. It frustrated some of the CIA station chiefs and case officers who worked closely with security services in these countries that their reports about the Islamists tended to be discounted in Washington.
Frank Anderson, the former Afghan task force director during the anti-Soviet covert war, had been promoted to run the Near East Division of the CIA’s Directorate of Operations, responsible for all espionage and covert action in South Asia and the Middle East. Anderson had been a strong advocate of the CIA’s support for Hekmatyar during the anti-Soviet jihad. He now argued that the returning jihadist veterans from Afghanistan were not as important as the Egyptian, Algerian, and Tunisian governments believed. Anderson argued that many Islamist radicals who claimed they had fought in Afghanistan were exaggerating their jihadist credentials. Any careful reading of Egyptian and Algerian history showed, Anderson argued, that radical Islam did not need to be imported from Afghanistan to fire a violent insurgency.7
All these fragmentary pieces of intelligence and hypotheses about the Afghan veterans swirled and recirculated by cable between Langley and the field. There was no consensus about what it all meant or how to respond. Still, in the CIA’s Counterterrorist Center and at the State Department’s Bureau of Intelligence and Research, a new analytical theory about Islamist terrorism gradually took hold.
Paul Pillar coined the phrase “ad hoc terrorists” to describe Ramzi Yousef and the World Trade Center plotters. While it was still possible that a government had a hand in the bombing, this seemed unlikely as the months passed. Yousef and his gang did not appear to belong to any formal group, despite their claims about a “Liberation Army, Fifth Battalion.” The Yousef plotters were clearly connected to international jihad support networks in Peshawar and the Middle East, but the extent and importance of these ties were unclear. Pillar later dropped the “ad hoc” term because he feared it seemed too casual, as if Yousef and his pals had been drinking coffee one afternoon and decided spontaneously to go bomb a building. But he and other senior analysts at the Counterterrorist Center persisted in their belief that the World Trade Center conspiracy marked a watershed in global terrorism, the debut of a new blend of unaffiliated mobile religious violence.
The CIA was slow to confront this new enemy, however, even after it had been identified. Agency analysts saw Iran’s intelligence service and proxy forces as a much graver terrorist threat to the United States than the Afghan veterans. Iranian-trained Hezbollah operatives bombed the Israeli cultural center in Argentina. Bitterness lingered at the CIA over Hezbollah’s 1984 torture and murder of Beirut station chief William Buckley.
To the extent they received government support, the new Islamists found money and guns not in Tehran but in Saudi Arabia. Yet the CIA and the White House were reluctant to confront the role of Saudi Arabian proselytizers, financiers, and government agencies. In Riyadh the CIA made little effort to recruit paid agents or collect intelligence about these threats. Diplomats at the State Department worried that because Saudi Arabia remained such a crucial security partner and oil supplier for the United States, the price of getting caught in an espionage operation in the kingdom might be unusually high. The CIA still ran intelligence collection operations in Saudi Arabia, but they tended to be cautious. They relied on technical intercepts more than penetration agents. They concentrated on traditional subjects such as succession and rivalry within the Saudi royal family.8
Saudi intelligence chief Prince Turki al-Faisal remained the CIA’s key liaison in the kingdom. He had an odd relationship with President Clinton, rooted in their time at Georgetown University. As he planned his run for the presidency from Little Rock, Clinton had gathered the addresses of his former classmates. He wrote them letters to ask for support. In his office at the General Intelligence Department in Riyadh, Prince Turki was surprised and entertained to receive one of these solicitations. At first he ignored it. He couldn’t remember Clinton from their university days, and he doubted the governor of a tiny state had much political future. As Clinton’s campaign gathered steam, Turki reevaluated. It could be useful for Saudi Arabia’s intelligence chief to have a personal connection with a prospective American president. He wrote to Clinton and opened a correspondence.
In the late spring of 1993, Georgetown University held a class reunion, and Prince Turki attended. Afterward the Saudi spy chief accompanied Frank Anderson from the CIA and Prince Bandar, the Saudi ambassador in Washington, to the White House. They sat down with Clinton and listened as he talked in meandering and general terms about globalization. The president’s discourse turned to the Middle East and Central Asia, and he asked Prince Turki: What policies should the United States pursue in countries such as Uzbekistan and Kazakhstan?
It was a typical Clinton session, more seminar than formal meeting. He was with some smart and interesting new people, and he wanted to hear their ideas about where American foreign policy should go. But Bandar and Turki left the White House disconcerted, shaking their heads. Clinton’s questions about how America should define its policies left the Saudis uneasy. They said to each other, He’s asking us?9
Still, shortly after the White House meeting, they sent a check for $20 million to fund a Middle Eastern studies program at the University of Arkansas, for which Clinton had tried to raise money while he was governor.10 It was a Saudi handshake, a small housewarming gift for a new friend.
KABUL PLUNGED INTO VIOLENCE and deprivation during 1993. Hekmatyar pounded the city indiscriminately with hundreds of rockets from his ample stores, killing and wounding thousands of civilians. The old mujahedin leaders realigned themselves in bizarre temporary partnerships. They fought artillery duels along Kabul’s avenues, dividing the city into a dense barricaded checkerboard of ethnic and ideological factions. Shiite militia fought against Hekmatyar around Kabul’s zoo, then switched sides and fought against Massoud. Sayyaf’s forces allied with his old Islamic law colleague Rabbani and hit the Shiites with unrestrained fury, beheading old men, women, children, and dogs. Dostum’s Uzbek militias carried out a campaign of rapes and executions on Kabul’s outskirts. Massoud hunkered down in the tattered defense ministry, a decaying former royal palace, and moved his troops north and south in running battles. The electricity in Kabul failed. The few remaining diplomats husbanded petrol for generators and held conferences by candlelight. Roads closed, food supplies shrank, and disease spread. About ten thousand Afghan civilians died violently by the year’s end.11
Prince Turki flew into Islamabad for meetings with the Afghan faction leaders, hired former Pakistani intelligence chief Hamid Gul as a mediation partner, and tried to talk the combatants into a settlement. They worked with the current ISI chief, Lieutenant General Javed Nasir, who wore a long beard and openly preached Islamist theology at mosques and public meetings. He was the most explicitly religious leader of Pakistani intelligence in a generation. Even some of Nasir’s colleagues within ISI were alarmed by his open proselytizing. They considered it a breach of the army’s professional traditions.12
Edmund McWilliams, the State Department diplomat who had campaigned from within the Islamabad embassy against the Islamist agendas of Pakistani and Saudi intelligence during the late 1980s, had recently been transferred to Central Asia. He watched the civil war with growing disgust. He sent a Confidential cable to Washington early in 1993 titled “Implications of Continued Stalemate in Afghanistan.” McWilliams argued that the “principled U.S. posture of letting Afghans find solutions to ‘their problems’ fails to take into account a central reality: Intense and continuing foreign involvement in Afghan affairs—by friendly and unfriendly governments and a myriad of well-financed fundamentalist organizations—thus far has precluded Afghans from finding ‘their own solutions.’ ” The hands-off policy of the United States “serves neither Afghan interests nor our own. . . . The absence of an effective Kabul government also has allowed Afghanistan to become a spawning ground for insurgency against legally constituted governments. Afghan-trained Islamic fundamentalist guerrillas directly threaten Tajikistan and are being dispatched to stir trouble in Middle Eastern, southwest Asian, and African states.”13
The McWilliams cable landed in a void. The White House formulated no policy toward Afghanistan during Clinton’s first term other than a vague endorsement of fitful, quixotic efforts by the United Nations to negotiate peace. This left American policy solely in the hands of the State Department, the agency that represented the United States in every country worldwide even when there was no policy to represent. Neither Warren Christopher nor any of his deputies had any interest in Afghanistan. Christopher said he intended to stand behind “the Americas desk,” meaning that foreign policy under Clinton would be managed to support domestic policies. Clinton appointed an acquaintance from his days at Oxford University as his assistant secretary of state for South Asia, in charge of diplomacy for India, Pakistan, and Afghanistan. Robin Raphel was a career foreign service officer who had risen to the rank of political counselor in the U.S. embassy in New Delhi, but she was relatively junior for the job. Blue-eyed, blond, and statuesque, she was an elegant, bright woman with an upper-crust air, and she was a serious equestrian. Apart from her personal history with the president, she had few connections at the White House or with the new team that had taken power at the State Department.14
Raphel tried to argue for continued humanitarian aid to Afghanistan. But as Clinton attempted to balance the federal budget after years of deficit spending, his administration drastically cut funding for the Agency for International Development, the government’s main overseas aid organization. Clinton directed available funds away from countries like Afghanistan and toward the neediest cases in Africa, a dying continent that Lake and the new AID director, Brian Atwood, felt had been neglected for too long by Republican administrations. “Nobody wanted to return to the hot spots of the Reagan-Bush years,” such as Afghanistan, recalled one member of Clinton’s team at the aid agency. “They just wanted them to go away.” South Asia was “just one of those black holes out there.” Atwood faced hostility from Republicans in Congress who argued that American development aid was being wasted in poor, chaotic countries. After heated arguments within AID and despite resistance from Raphel, the United States ended all bilateral development aid to Afghanistan less than two years after Clinton took office.15
CIA director James Woolsey saw Afghanistan in these months merely as “a place where there was a lot of warlord-ism.” Its civil war and its jihad training camps did not seem to him a significant factor in the rise of Islamist politics in North Africa. In this analysis he was influenced by Frank Anderson, his Near East operations chief. Woolsey liked and admired Anderson and relied on him heavily for analysis about Afghanistan and the Arab world.16
The CIA was active in Central Asia. After the Soviet Union’s collapse the CIA’s Directorate of Operations moved into the newly independent, former Soviet republics. Among other objectives the CIA sought to thwart Iranian ambitions in Central Asia. Officers tracked Iranian agents and tried to secure the region’s loose nuclear bombs and materials. Oil-rich republics along the Caspian Sea opened their vast energy reserves to foreign corporations. American firms sought a piece of the action. For all these reasons the CIA’s officers “were all over Ukraine and Central Asia, going in just as fast as we could, finding new opportunities,” recalled Thomas Twetten, then chief of the Directorate of Operations and Frank Anderson’s supervisor.17
But the CIA ignored Afghanistan and its civil war. Twetten felt there was nothing the United States could do to mediate the Afghan conflict or put the country back together again. There were too many other challenges in a world so suddenly and vastly changed by communism’s collapse. The Afghan war threatened to destabilize the new Central Asian countries, but even that danger seemed remote. Afghanistan was “just really background” at the Directorate of Operations only two years after it had been at the center of one of the CIA’s most important and richly funded covert programs.18
Charles Cogan, the former Near East Division chief who had helped create the anti-Soviet jihad, spoke for many at the agency during this period when he described the CIA and the Islamist rebels built up by Pakistani intelligence as merely “partners in time.” They had no enduring interests in common. The United States “was no more able to put together a polity in the ghost town that Kabul had become than it can in Dushanbe or, alas, in Mogadishu. Nor should it try.” American intervention in civil wars unfolding in former Cold War client states would “lead to a dangerous overextension of American forces and resources [and] it will draw upon us more hatreds and jealousies. In most cases, we would be well advised to maintain a prudent distance, in the words of Douglas MacArthur, from the ‘internal purification problems’ of others.”19
Afghanistan was indeed about to purify itself. It was about to disgorge a radical Islamic militia as pure and unbending in its belief system as any in the Muslim world since King Saud’s antimodern Ikhwan had stormed across the Arabian peninsula seven decades before.