Iran's Nukes: What US Intelligence Really Believes

By James Risen, The New York Times

18 March 12

hile American spy agencies have believed that the Iranians halted efforts to build a nuclear bomb back in 2003, the difficulty in assessing the government's ambitions was evident two years ago, when what appeared to be alarming new intelligence emerged, according to current and former United States officials.

Intercepted communications of Iranian officials discussing their nuclear program raised concerns that the country's leaders had decided to revive efforts to develop a weapon, intelligence officials said.

That, along with a stream of other information, set off an intensive review and delayed publication of the 2010 National Intelligence Estimate, a classified report reflecting the consensus of analysts from 16 agencies. But in the end, they deemed the intercepts and other evidence unpersuasive, and they stuck to their longstanding conclusion.

The intelligence crisis that erupted in 2010, which has not been previously disclosed, only underscores how central that assessment has become to matters of war and peace.

Today, as suspicions about Iran's nuclear ambitions have provoked tough sanctions and threats of military confrontation, top administration officials have said that Iran still has not decided to pursue a weapon, reflecting the intelligence community's secret analysis. But if that assessment changes, it could lift a brake set by President Obama, who has not ruled out military options as a last resort to prevent Iran gaining nuclear arms.

Publicly and privately, American intelligence officials express confidence in the spy agencies' assertions. Still, some acknowledge significant intelligence gaps in understanding the intentions of Iran's leaders and whether they would approve the crucial steps toward engineering a bomb, the most covert aspect of one of the most difficult intelligence collection targets in the world.

Much of what analysts sift through are shards of information that are ambiguous or incomplete, sometimes not up to date, and that typically offer more insight about what the Iranians are not doing than evidence of what they are up to.

As a result, officials caution that they cannot offer certainty. "I'd say that I have about 75 percent confidence in the assessment that they haven't restarted the program," said one former senior intelligence official.

Another former intelligence official said: "Iran is the hardest intelligence target there is. It is harder by far than North Korea.

"In large part, that's because their system is so confusing," he said, which "has the effect of making it difficult to determine who speaks authoritatively on what."

And, he added, "We're not on the ground, and not having our people on the ground to catch nuance is a problem."

Iran says its nuclear program is for peaceful civilian purposes, but American intelligence agencies and the International Atomic Energy Agency have picked up evidence in recent years that some Iranian research activities that may be weapons-related have continued since 2003, officials said. That information has not been significant enough for the spy agencies to alter their view that the weapons program has not been restarted.

Mossad, Israel's intelligence service, agrees with the American intelligence assessments, even while Israeli political leaders have been pushing for quick, aggressive action to block Iran from becoming what they describe as an existential threat to the Jewish state.

"Their people ask very hard questions, but Mossad does not disagree with the U.S. on the weapons program," said one former senior American intelligence official, who, like others for this article, would speak only on the condition of anonymity about classified information. "There is not a lot of dispute between the U.S. and Israeli intelligence communities on the facts."

In trying to evaluate the potential perils of Iran's nuclear program, the United States' spy agencies have spent years trying to track its efforts to enrich uranium and develop missile technology, and watching for any move toward weaponization - designing and building a bomb.

Hunting for signs of the resumption of a weapons program is more difficult than monitoring enrichment and missile-building activities, both of which require large investments in plants, equipment and related infrastructure. American intelligence officials said that the conversations of only a dozen or so top Iranian officials and scientists would be worth monitoring in order to determine whether the weapons program had been restarted, because decision-making on nuclear matters is so highly compartmentalized in Iran.

"Reactors are easier to track than enrichment facilities, but obviously anything that involves a lot of construction is easier to track than scientific and intellectual work," said Jeffrey T. Richelson, the author of "Spying on the Bomb," a history of American nuclear intelligence. "At certain stages, it is very hard to track the weapons work unless someone is blabbing and their communications can be intercepted."

The extent of the evidence the spy agencies have collected is unclear because most of their findings are classified, but intelligence officials say they have been throwing everything they have at the Iranian program.

While the National Security Agency eavesdrops on telephone conversations of Iranian officials and conducts other forms of electronic surveillance, the National Geospatial-Intelligence Agency analyzes radar imagery and digital images of nuclear sites. Outside analysts believe high-tech drones prowl overhead; one came down late last year deep inside Iranian territory, though American officials said they lost control of it in Afghanistan.

Meanwhile, clandestine ground sensors, which can detect electromagnetic signals or radioactive emissions that could be linked to covert nuclear activity, are placed near suspect Iranian facilities. The United States also relies heavily on information gathered by inspectors with the International Atomic Energy Agency who visit some of Iran's nuclear-related facilities.

But collecting independent human intelligence - recruiting spies - has been by far the most difficult task for American intelligence. Some operational lapses - and the lack of an embassy as a base of operations ever since the hostage crisis three decades ago - have frequently left the C.I.A. virtually blind on the ground in Iran, according to former intelligence officials.

In 2004, for example, the C.I.A. put a whole network of Iranian agents in jeopardy after a technological mistake by an agency officer, according to former intelligence officials.

In 2005, a presidential commission that reviewed the prewar failures of the intelligence on Iraq's supposed weapons programs faulted American intelligence on Iran, saying it included little valuable information from spies.

More recently, the C.I.A. suffered a setback in efforts to question Iranian exiles and recruit nuclear scientists. Two years ago, agency officials had to sort through the wreckage of the strange case of Shahram Amiri, an Iranian scientist who apparently defected to the United States in 2009 and then returned to Iran in 2010 after claiming he had been abducted by the C.I.A.

His case is eerily similar to that of Vitaly Yurchenko, a K.G.B. officer who defected to the United States in 1985 and went back to the Soviet Union later that year, claiming he had been drugged and kidnapped by the C.I.A.

Like Mr. Yurchenko, Mr. Amiri's case has provoked debate within the agency about whether he was a genuine defector, and whether any of the information he provided could be trusted.

The United States and Israel share intelligence on Iran, American officials said. For its spying efforts, Israel relies in part on an Iranian exile group that is labeled a terrorist organization by the United States, the Mujahedeen Khalq, or M.E.K., which is based in Iraq. The Israelis have also developed close ties in the semiautonomous region of Kurdistan in northern Iraq, and they are believed to use Kurdish agents who can move back and forth across the border into Iran.

American intelligence officials, however, are wary of relying on information from an opposition group like the M.E.K., particularly after their experience in Iraq of relying on flawed information provided by the Iraqi National Congress, an exile group run by Ahmad Chalabi.

"I'm very suspicious of anything that the M.E.K. provides," said David A. Kay, who led the C.I.A.'s fruitless effort to find weapons program in Iraq. "We all dealt with the Chalabis of the world once."

Just as in 2010, new evidence about the Iranian nuclear program delayed the National Intelligence Estimate in 2007, the last previous assessment. Current and former American officials say that a draft version of the assessment had been completed when the United States began to collect surprising intelligence suggesting that Iran had suspended its weapons program and disbanded its weapons team four years earlier.

The draft version had concluded that the Iranians were still trying to build a bomb, the same finding of a 2005 assessment. But as they scrutinized the new intelligence from several sources, including intercepted communications in which Iranian officials were heard complaining to one another about stopping the program, the American intelligence officials decided they had to change course, officials said. While enrichment activities continued, the evidence that Iran had halted its weapons program in 2003 at the direction of the supreme leader, Ayatollah Ali Khamenei, was too strong to ignore, they said.

One former senior official characterized the information as very persuasive. "I had high confidence in it," he said. "There was tremendous evidence that the program had been halted."

And today, despite criticism of that assessment from some outside observers and hawkish politicians, American intelligence analysts still believe that the Iranians have not gotten the go-ahead from Ayatollah Khamenei to revive the program.

"That assessment," said one American official, "holds up really well."