In January 2007, Israeli intelligence officials were horrified by information acquired when Mossad agents broke into the hotel room of a senior Syrian official in London and downloaded the contents of his laptop. The pilfered files revealed that Syria, aided by North Korea, was building a nuclear reactor that could produce an atomic bomb.
Until then, according to military intelligence officials, Israeli intelligence thought Syria had no nuclear program. But that was because Syria’s president, Bashar al-Assad, had set up a parallel and separate system of command and control for building the reactor. The discovery caused a panic in Israel, and grave concern in Washington, which had relied heavily on Israel’s assurances that it knew everything about Syria. By the time the reactor was discovered, it was almost ready to become operational. Deeming it too late for diplomatic action, Israel asked the Americans to attack the facility that spring. President George W. Bush refused, so Israel went ahead on its own, destroying the reactor in a bombing raid on Sept. 6, 2007, and risking a war.
The lesson in humility that intelligence agencies should have learned from this affair is evident. I was therefore surprised by the assertions of American and Israeli officials I interviewed while researching a recent article on the prospect of an Israeli attack against Iran. They repeatedly stated, “we will know,” when we talked about the possibility of Iran’s moving to produce nuclear weapons.
This month, playing down reports of disagreements between the United States and Israel, a senior American official said it plainly. “There is day-to-day intelligence cooperation between the United States and Israel,” he told me. “If and when the Iranians decide to go over to nuclear weapons production,” he added, “we’ll know about it and we’ll share the information between us.”
Today, the two countries agree that Iranians have not begun to assemble a nuclear device, won’t do so until their supreme leader gives them a go-ahead, and would need about nine months to create a bomb. Both countries are convinced that they will obtain unequivocal intelligence when the order to start is given.
Here and in Washington, officials assume that as soon as information is received that Iran has moved to build nuclear weapons, Israel will decide to attack its nuclear facilities. Therefore, any exchange of intelligence between Israel and the United States could have far-reaching consequences for the Middle East — and the world.
In other words, the momentous decision will be driven to an extraordinary extent by intelligence reports. But even though intense focus on Iran’s nuclear program has presumably increased the volume of intelligence gathered about it, it remains true that intelligence officers tend to rely heavily on a few trusted sources. And it may be only human for a case officer to be excited by discovering something that appears to be a secret.
Bits of data can be misread, however, and erroneous analysis has a habit of finding its way to those most eager to use it. So in watching Iran’s nuclear project, even a slight intelligence gaffe could have an outcome of historic proportions.
In the late 1980s, the United States and Israel believed that they had good intelligence on Iraq, but they missed the extent of Saddam Hussein’s pursuit of unconventional weapons — until after he invaded Kuwait. Like the Syrian nuclear project, that stands as a warning of the dangers of overconfidence that a threat is absent. A parallel case in Iran might arise if the supreme leader were to convey his order to assemble a nuclear bomb via channels the C.I.A. and Mossad knew nothing about.
A surprise could also come from the opposite direction — as a result of overestimating the other side’s capabilities and intentions; in Iran’s case, that could lead to a premature Israeli attack.
Could that happen? It did — to America when it relied on faulty intelligence to justify its invasion of Iraq in 2003, claiming that it knew Saddam Hussein was hiding unconventional weapons.
This year, an equally fateful decision may well rely on the quality of available intelligence. So, caution is in order: Relying on intelligence as the chief touchstone for decisions about whether and when to attack creates a wide opening for misunderstandings, divergent interpretations and vulnerabilities to parties with an interest in either attack or delay.
Both Israel and America should acknowledge that scraps of information cannot serve as the basis for action against Iran, and they should find new criteria for such a decision.
The direct talks that begin on Friday in Istanbul offer a chance to develop at least some new measures of Iran’s willingness to cooperate. This could work on two levels. Publicly, Iran can be judged on its willingness to meet the world’s demands. Under the surface, a continuing dialogue between Iran and the West — which will require much consultation and interaction among Iranians themselves — can reveal or clarify information about Iran and its decision making that would not have surfaced otherwise.
Even without firm knowledge that the supreme leader had ordered his scientists to assemble a bomb, such information could help guide analysis and decision making, which nobody can afford to approach lightly. A miscalculation could be the worst possible outcome.
Ronen Bergman, a senior political and military analyst for Yediot Aharonot, is at work on a history of the Mossad.