Dept Of Secrets

Israel's Secret Uranium Buy

How Argentina fueled Ben-Gurion's nuclear program.

In mid-July 1964, the State Department and the CIA sent a joint message asking the U.S. embassies in Argentina and Israel to check out an unverified intelligence report. They wanted to know whether the Argentines had agreed to sell Israel some 80-100 tons of uranium oxide, or "yellowcake," an essential product for fueling a nuclear reactor and thereby producing plutonium that can be used in weapons.

As it turned out, Washington had gotten information about the sale from the British government, which in turn had found out about it from the Canadians. All three governments were concerned about Israel's nuclear weapons ambitions, and the yellowcake transaction was strong evidence that something was amiss. American diplomats in Argentina confirmed the sale, which soon put the State Department in an awkward position: It would have to ask the Israelis about a transaction that flew in the face of their assurances that the country's nuclear program was for peaceful purposes only.

Israel's nuclear program presents a sort of paradox to historians. While it may be the world's worst-kept secret, it is also the world's most opaque nuclear program. One aspect of the Israeli nuclear program that has been especially mysterious is how and where Israel was able to obtain the raw material required to sustain a serious weapons effort. In the 1960s, this was a real challenge for U.S. intelligence, which was not entirely clear about the purposes of the Israeli program or whether Israel would comply with its "peaceful use only" pledge. It remains a challenge for historians today because Israel still does not acknowledge that it has nuclear weapons.

Previously obscure declassified archival documents on the yellowcake sale shed light on the global context of the Israeli nuclear program's early history. Edited and annotated by the two of us, 42 documents are being published today for the first time by the National Security Archive and the Nuclear Proliferation International History Project (in conjunction with the James Martin Center for Nonproliferation Studies). They demonstrate how vigorously Israel sought raw materials for its nuclear program and how persistently it tried to cultivate relations with nuclear suppliers. They also tell us how other players -- in particular the United States, Britain, and Canada -- viewed the program.

The story of the Argentine yellowcake sale to Israel has remained largely untold because Israel has gone to great lengths to keep it secret and because the U.S. government and its close allies have kept quiet about what they knew at the time. The United States has always been ambivalent about Israel's nuclear program, and exposing what it knew or suspected about the Israeli nuclear program could have caused the United States serious diplomatic problems with Israel's Arab neighbors and possibly the Soviet Union. This limited what Washington could do to circumscribe Israel's nuclear ambitions. Any kind of serious economic or political pressures, even if only contemplated by some, would have become public. And that could have been explosive domestically and internationally.

The U.S. government had been worried about an Israeli nuclear weapons program since late-1960, when the CIA learned and confirmed that, for nearly two years, Israel had been constructing a major nuclear facility (a reactor plus related infrastructure), with French assistance, near the town of Dimona in the Negev Desert. Initially, in support of Israel's desire to establish a nuclear program with military potential, the French apparently agreed to provide Israel with reactor fuel under loose safeguards. Under Charles De Gaulle, however, French policy changed, and it appears that by 1963, when the reactor was near completion, France imposed major constraints on supplying uranium to Dimona.

The Israelis had been trying to extract uranium from phosphate but that proved too costly; they needed a source that they could use freely, without external safeguards. South Africa was a prospective source. The French themselves recognized that Israel could try to acquire uranium from other countries, such as Argentina or Belgium, and in early 1964 they asked Washington whether the Israelis had "tapped" any such sources.

The Canadian government was interested in the Israeli nuclear program from its very inception. When Israeli Prime Minister David Ben-Gurion met Prime Minister John Diefenbaker on May 25, 1961, Dimona was at the center of the discussion. As he did with President John Kennedy a few days later, Ben-Gurion pledged that the Dimona project was peaceful. In March 1964, Canadian intelligence analyst Jacob Koop prepared a long secret report on Israel's nuclear program, asserting that Israel had all of the "prerequisites for commencing a modest nuclear weapons development project."

Not long after this report was prepared, Canadian intelligence learned (from a still-unknown source) that the Argentine government had made arrangements to supply 80-100 tons of yellowcake to Israel. By the end of April 1964, the British had seen the Canadian report. According to a British diplomat, "This means that Israel now has virtually unlimited supplies of uranium free of safeguards." Moreover, if the Israelis had reprocessing facilities, they could produce enough plutonium to "fuel a nuclear bomb" 18-20 months from the beginning of 1964.

The British soon shared the Canadian report with U.S. intelligence, overcoming Canadian reluctance to share it with its neighbors to the south (apparently the Canadians were irritated that the United States would not share the results of a recent American visit to Dimona). The CIA was initially skeptical, but in June 1964, the State Department and the CIA decided that the story should be checked out and sent the query -- reproduced below -- to its embassies in both Argentina and Israel. In September, the U.S. embassy in Buenos Aires confirmed from local sources that during 1963 Israel had arranged to purchase 80 tons of yellowcake from Argentina.

Evidently, the United States took seriously the information it obtained about the Argentine yellowcake sale. Like its British and Canadian allies, Washington was concerned that an Israeli bomb would threaten stability in the Middle East and complicate American efforts to curb nuclear proliferation worldwide. Moreover, to ensure that the Israelis were abiding by their public pledge that the Dimona facility was for "peaceful" use only, Kennedy and Prime Minister Levi Eshkol had secretly agreed in the summer of 1963 to allow American scientists to visit the reactor. The first U.S. team arrived in Dimona in early January 1964, but it is now known that the Israelis made "special arrangements" to prevent the visitors from seeing anything that revealed the true nature of the project.

In the fall of 1964, not long after the yellowcake sale was confirmed, U.S. diplomats brought the matter up with Argentine officials. While they did not object to the sale, they were concerned that there were no safeguards beyond a general agreement on peaceful purposes. The State Department wanted the International Atomic Energy Agency (IAEA) to be allowed to supervise future sales through reports and inspections. Admiral Oscar A. Quihillalt, the director of Argentina's atomic energy program, appeared sympathetic to U.S. concerns but said he could not do anything to address them. The sale to Israel could not be reversed or changed.

The U.S. embassy in Israel, including the CIA station, could learn nothing about the yellowcake from local sources, so the State Department asked Ambassador Walworth Barbour to go to a higher level. In June 1966, Barbour spoke directly with Foreign Minister Abba Eban. He was instructed to say that the yellowcake was "precisely [the] type of problem which augments apprehension" at "high levels" in Washington and elsewhere about nuclear proliferation. The problem also illustrated the "need for IAEA safeguards to reassure world of peaceful orientation [of] Israeli nuclear program." Barbour met Eban several times, but the latter was evasive -- apparently because he was not in the loop -- saying that he would confer with the deputy defense minister, Zvi Dinstein, who "keeps the store." If Eban ever provided Barbour with an answer, it has not surfaced in the archival record. Despite Israel's evasions, Washington apparently took no counteraction, but only continued to keep tabs through visits to the Dimona plant.

While the U.S. government was exploring the Argentine sale, it also investigated rumors during the spring of 1965 that the French uranium company in Gabon had asked Paris for permission to sell yellowcake to Israel. The French had already stopped such an effort in 1963, but when U.S. embassy officials in Gabon asked company officials about the rumored sales, no one would give any answers. As the French government controlled the exports, it was unclear whether the Gabonese or local company officials could actually divert uranium. Whether Israel received any yellowcake from Gabon during the 1960s still remains a mystery. In any event, sometime in mid-1968, Israel acquired 200 tons of yellowcake from Belgium in a complex clandestine operation known as the "Plumbat" affair, which involved a Mossad-run Italian front company and the at-sea transfer of uranium from a European cargo ship to an Israeli freighter.

The yellowcake issue was an important Israeli nuclear secret, but its biggest nuclear secret was the existence of a reprocessing facility to transform spent reactor fuel from Dimona into weapons-grade plutonium. For example, according to an October 1964 Special National Intelligence Estimate on nuclear proliferation, a "major deficiency, in terms of a weapons program, is the lack of a plutonium separation plant." The Israelis had told the Canadians and the Americans in 1961 that Dimona would include a pilot plant for reprocessing, but it was presumed that it would be too small to support a weapons program. In reality, however, the original French design for Dimona included a large underground reprocessing facility; this was Israel's most important nuclear secret, which Dimona technician Mordecai Vanunu later made public. Today, it is unclear exactly how much Western intelligence knew about the reprocessing facility and when and how it learned it.

The story of the yellowcake sale and the vain effort to prevent its military diversion is historical evidence of how difficult it was for the United States to stop Israel or any similarly determined government from undertaking a nuclear weapons program. That Israel was a major U.S. ally complicated matters. Tight secrecy about the Israeli nuclear program made it nearly impossible to raise pressure on Israel without risking an international incident.

From today's perspective, the story of the Argentine yellowcake highlights the continued lack of sufficiently tight international regulations for the trade of yellowcake. Despite U.S. support for tighter verification requirements during negotiation of the nuclear Nonproliferation Treaty in the 1960s, the agreements that the IAEA has with non-nuclear weapons states still do not require safeguards on the sale of yellowcake, only documentation of transfers. Changing this would be extremely difficult. That yellowcake remains politically charged was evident in the spurious allegations about Niger during the lead-up to the Iraq War. Secret controversies may still persist over who is selling yellowcake to whom and under what conditions, but we may never know.

Wikimedia Commons

National Security

Mission: Assassination

The CIA's a lot better at targeted killing now than it used to be.

In recent months, the public debate over targeted killings of terrorist suspects by the CIA and Joint Special Operations Command has been ratcheted up by reports of rampant drone strikes in Pakistan and Yemen, and by the Obama administration's attempts to keep under wraps its legal justifications for these operations. As it happens, debates over whether it's okay for the CIA to kill foreign nationals (leaving aside U.S. citizens), and under what circumstances, have been going on for years. The article below, from the CIA's in-house journal, Studies in Intelligence (obtained through the Freedom of Information Act by Jeffrey Richelson), traces the history of the issue back to the Ford administration's 1976 executive order banning "political" assassinations. According to the 1996 article, written by a CIA attorney, the agency began contemplating the "lethal use of force" almost from the time it was formed (in 1947). And as we all know, "at various times" over the next 30 years agency officials made numerous plans actually to assassinate foreign leaders -- Patrice Lumumba and Fidel Castro being the best-known targets. For better or worse, precisely none of these attempts succeeded, the author states.

The political landscape changed dramatically in the mid-1970s when Congress began uncovering rampant abuses by the intelligence community and instituting reforms and limits on the activities of the CIA, among others. Surprisingly, Congress could not make up its mind to put a stop to assassinations, leaving it up to President Gerald Ford to do so. The administration's thinking, the author says, was shaped by a mix of moral and practical considerations -- not least being the worry that U.S. leaders might themselves become targets in retaliation. Jimmy Carter and Ronald Reagan signed their own executive orders continuing the prohibition. Carter broadened it by dropping the word "political," and Reagan agreed with that formulation. 

But for those hoping the matter might be laid to rest right there -- including reportedly many inside the CIA -- big questions remained. First of all, even though the basic objective of these presidential orders was supposedly well understood, according to the article, it was not entirely clear what key terms meant. Even after Carter dropped the word "political," the new wording "continues to engender discussion" within the key agencies involved because "the parameters of simple 'assassination' are not always clear." Beyond that, the article says, "the United States [has] retained the options of encouraging coups" and other potentially violent activities, and "the President still may authorize CIA to conduct operations abroad that endanger the lives of others."

The trick, apparently, for those engaged in approving, carrying out, or reviewing these vaguely defined activities, has been how to strike a balance between achieving real-world anti-terrorist or similar objectives while staying within the law. The article lays out scenarios that bring the issue into relief, including going into detail about the conundrum of how to deal with Panama's Manuel Noriega in the late 1980s. Particularly telling are a couple of hilarious Gary Trudeau cartoons the author includes in order to drive home the point that trying to put legal handcuffs on bare-knuckled covert operations can quickly appear ludicrous.

Written back in the Clinton era, the Studies in Intelligence article may seem somewhat out of date. As a recent New York Times piece on the same issue noted, after the 2001 terrorist attacks, any internal concerns over CIA involvement with targeted killings "were quickly swept aside." But at least one major fact has not changed -- the only formal constraint that exists against assassinations by the CIA is not the law but a mere presidential order, which the commander-in-chief can, in theory, easily revoke.

David Burnett/Newsmakers