In the West the weaknesses of the Non-Proliferation Treaty were understood from the start. For the treaty to have weight it would have to be backed by the threat of sanctions—but even then, given the willingness of governments to "eat grass" to acquire such military capabilities, it was unlikely to deter serious aspirants from pursuing the bomb. The solution, therefore, would lie in the complex realm of export controls—restrictions on the sale of nuclear-related materials and components that might appear to be for peaceful purposes (research, health care, power generation) but could be used for weapons development. Emphasis was to be put on technologies that would allow countries to become self-sufficient in nuclear fuels—on uranium-enrichment and plutonium-extraction plants. Exports would be allowed to countries that had joined the treaty, subject to IAEA scrutiny on the ground, but would be banned to countries that had refused to sign, like Pakistan. The reliance on the United Nations posed obvious operational problems: the IAEA was a politicized bureaucracy, awash in national jealousies, and staffed by functionaries who considered themselves to be in the business primarily of encouraging nuclear-energy development. Nonetheless, in the early and mid-1970s two groups of technologically advanced countries (diplomatic assemblies known as the Zangger Committee and the Nuclear Suppliers Group) began to meet to decide on the lists of restricted materials and equipment and to negotiate the tricky terrain of national implementation and cooperation between participating governments. Over the thirty ensuing years their record has been mixed. Though they have produced ever longer export control lists that have helped to slow the nuclear trade by forcing more of it underground, they themselves have been stymied by national bureaucracies, slowed by governmental reluctance to interfere with lucrative business deals, and frustrated by the depths of global trade. As a result their lists have lagged behind the market they intend to regulate. And at no point have they been a match for energies like those of A. Q. Khan.
In fairness, Khan was an extraordinarily aggressive man. After his return to Pakistan, in December of 1975, he spent a few months within the confines of the Pakistan Atomic Energy Commission, but he was frustrated by its slow pace, and he angrily burst free. In a private conversation with Bhutto he accused the PAEC chairman—the despised Munir Ahmed Khan—of betraying the country. As he later remembered the conversation to Zahid Malik, Khan said to Bhutto, "Munir Ahmed Khan and his people are liars and cheats. They have no love for the country. They are not even faithful to you. They have told you a pack of lies. No work is being carried out, and Munir Ahmed Khan is cheating you." What he did not say, but at some point apparently believed, was that Munir Ahmed Khan had been turned by his tenure at the IAEA, and was actively subverting Pakistan's nuclear goals. Evidently Bhutto was too shrewd to be convinced, because he never took action against A. Q. Khan's enemies; but with nothing to be gained by frustrating Khan, and with perhaps some benefits to accrue from setting up a competition with the PAEC, he decided to give Khan full autonomy, and promised him a large and secret budget. He must have thought he was doing Khan a favor.
On July 31, 1976, Khan founded the Engineering Research Laboratories to set up and operate a centrifuge plant based on the stolen urenco designs. The raw uranium was to be mined in central Pakistan, converted to gas, and sent to Khan to spin very fast. If anyone asked, the stated purpose was to produce uranium enriched merely to the level necessary for electric-power generation—albeit completely outside of IAEA verification and controls. The plant was to be built as a complex of industrial-looking buildings among low hills about forty miles southeast of Islamabad, in an out-of-the-way town called Kahuta, which could be locked down and guarded by Pakistani security forces. Because Khan felt that his nation's survival was at stake, he proceeded not sequentially but simultaneously on multiple fronts—hiring staff, laying out the installations, initiating the construction of the Kahuta plant, and setting up a pilot project elsewhere to resolve the practical intricacies of linked centrifuges and to make the first trial runs. This was a big operation. Ultimately he hired as many as 10,000 people. Most important, he launched the procurement effort in Europe and the United States.
The U.S. government knew very well what was happening; Bhutto had made no secret of his ambitions, and by conventional logic it made sense for Pakistan to acquire a nuclear bomb. As an element of Cold War strategy, Pakistan remained a U.S. client state, somewhat prickly under Bhutto, but supported by American aid, and still quite accessible to American diplomats and officials. It is reasonable to assume—and was always presumed within Khan's inner circle to be true—that the CIA had penetrated both the PAEC and Kahuta from an early date. Given the size of the programs under way, this would have been easy to do. The view from the inside was sobering: despite an assumption among European governments that Pakistan lacked the necessary technical expertise, it became clear that this effort was serious, and that it was likely to succeed. Such an outcome seemed all the more worrisome in Washington, D.C., because Bhutto had resentfully mentioned Christian, Jewish, Hindu, and Communist bombs, and the possibility therefore existed that a Pakistani device would amount to more than a counterbalance to India—that it would be handled as a "Muslim" bomb to be spread around. Apparently other countries had the same idea, though with hope rather than fear: Libya and Saudi Arabia, for example, are both suspected of having funded Khan early on, probably with the expectation of a return. In any case, by the late 1970s, as Khan proceeded determinedly and American appeals to desist were rebuffed by the government in Islamabad, U.S. officials realized that the only chance they had to stop Pakistan from building a bomb was to take the supply-side approach—to block Pakistan's procurements abroad.
Blocking procurements within the United States proved to be relatively easy, because Khan had few American contacts, and U.S. export-control lists were already quite extensive—significantly more so than those that had been agreed upon by the international supplier groups. Moreover, deep within the customs and commerce bureaucracies, where such regulations are effectuated (or not) day to day, American officials, as representatives of a dominant nuclear power, tended naturally to agree on the importance of nonproliferation, and were alert to hints of violations that appeared in the paperwork that crossed their desks. As a result, though some transactions slipped by unseen, the U.S. government thwarted most of the attempted acquisitions from American suppliers.
The export-control record was altogether different in Europe, where constellations of companies were selling their wares to the Pakistanis, often with the tacit or explicit approval of their governments. In a breathless but generally reliable book titled The Islamic Bomb, published in 1981, the reporters Steve Weissman and Herbert Krosney tell a typical story of three of Khan's purchasing agents, who in 1976 went to a small Swiss company in a small Swiss town and proposed to buy its specialized high-vacuum valves for the express purpose of equipping a Pakistani centrifuge enrichment plant. The company dutifully checked with the Swiss authorities, who sent back a printout of their export regulations, including the list of restricted items as defined by the Nuclear Suppliers Group. Weissman and Krosney write,
Complete centrifuge units were listed, and could only be exported to [IAEA] safeguarded facilities, which the Pakistani enrichment plant was not. High-vacuum valves were not listed, even if expressly intended for a centrifuge enrichment unit. The valves might be necessary to the centrifuge. But, in the logic of the ... list, they were not "nuclear sensitive," and did not directly separate the two different uranium isotopes, uranium 235 and uranium 238.
The company, in other words, was informed that it could proceed with the sale, and so it did—as did others throughout Western Europe. In Holland, also in 1976, a Dutch company in the automotive-transmission business sold 6,500 high-strength steel tubes to Pakistan—tubes that could serve as the basic components of centrifuges. The Dutch government knew of the deal and advised against it, but the company sent their product anyway (initially claiming that the tubes were for agriculture), and argued that no export license was required by Dutch law. The argument was accepted, and further shipments went through without delay. Ultimately there were several paltry prosecutions, including one that led to the conviction of a Dutch businessman named Henk Slebos for illegally exporting an American-made Tektronix oscilloscope in 1983. Slebos was a personal friend of Khan's, and one of his main European suppliers. He was sentenced to a year in prison, but never served the time, and continued brazenly to send equipment to Pakistan. Controls were so loose that for more than a decade Khan himself kept visiting Europe.
Such was the scene American officials faced in the global nuclear marketplace as they grappled with the inadequacy of the UN's multi-party approach, and tried through private entreaties to European governments to prevent the spread of nuclear weapons. They were undercut, as they are today, by the thousands of nuclear warheads that the United States insisted on retaining for itself, and the resentment that such an obvious double standard provoked even within countries such as Germany and the Netherlands, which were said to be direct beneficiaries of American nuclear strength. They did, however, experience a few successes—particularly in 1977, when they pressured the French into backing out of the lucrative agreement to provide Pakistan with its long-desired plutonium-reprocessing plant. The cancellation set back the PAEC's nuclear-weapons plans by a decade or more. In consequence it further legitimized A. Q. Khan, and helped him to pursue his alternative goals—but nothing could be done about that anyway. For France the cost of killing the deal was several billion dollars, because of the loss of associated contracts for French products such as airplanes and trucks. The decision was all the more difficult because, with its "force de frappe," France embodied the right (and perhaps the need) of independent nations to bear nuclear arms. Such was its ambivalence that it had refused to join the Non-Proliferation Treaty. (It would not join until 1992.) Nonetheless, as an established power pretending to diplomatic relevance, it had little choice but to back away once it was faced with evidence of Pakistan's ambitions. By American estimation France this time behaved well.