WASHINGTON >> At the nation’s top spy agency, the ghosts of Iraq are never far away.

One CIA analyst who had helped develop some of the intelligence about Saddam Hussein’s supposed weapons of mass destruction had a breakdown months after the Iraq war began; he had participated in the post-invasion hunt there that found the weapons did not exist. When he eventually was given a new assignment assessing Iran’s nuclear program, he confided a fear to colleagues: that the intelligence community might get it wrong again.

“He felt enormous guilt that he had gotten us into the war,” said one former official who worked with the analyst. “He was afraid it was going to be deja vu all over again.”

Today, analysts and others at the CIA who are struggling to understand the nuclear ambitions of Iran are keenly aware that the agency’s credibility is again on the line, amid threats of new military interventions. The intelligence debacle on Iraq has deeply influenced the way they do their work, with new safeguards intended to force analysts to be more skeptical in evaluating evidence and more cautious in drawing conclusions.

Former intelligence officials say that this shows appropriate vigilance in dealing with often murky information, while some detractors argue that the agency is not just careful but overly skittish on Iran, reluctant to be blamed for any findings that might lead the United States to bloodshed.

“For a lot of people in the intelligence community, there is a feeling that they don’t want to repeat the same mistake,” said Greg Thielmann, a former State Department intelligence analyst who resigned to protest what he considered the Bush administration’s politicization of the prewar Iraq intelligence. “The intelligence community as a whole has better practices now partly because of the scar tissue they still have from Iraq,” added Thielmann, now a senior fellow at the Arms Control Association in Washington.

Paul Pillar, a former senior CIA analyst on the Middle East, says he believes that analysts are guided by the facts in making their assessments about Iran, but that they almost certainly have Iraq weighing on them.

“Because intelligence officials are human beings, one cannot rule out the possibility of the tendency to overcompensate for past errors,” said Pillar, now the director of graduate studies at the Security Studies Program at Georgetown University.

Top intelligence officials have said that analysts believe that Iran has been moving to expand its infrastructure and technological ability to become a nuclear power, but that the Iranian leadership has not made a decision to build an atomic bomb.

Current and former senior U.S. officials acknowledge, though, that there are significant gaps in their knowledge, and that they may not be able to quickly detect any decision to restart Iran’s weapons program, which they concluded had been halted in 2003.

After the misjudgments on Iraq, the CIA and other intelligence agencies imposed new checks and balances, including a requirement that analytical work be subjected to “red teaming.” That means a group of analysts would challenge the conclusions of their colleagues, looking for weaknesses or errors.

The intelligence community also now requires that analysts be told much more about the sources of the information they receive from the United States’ human and technological spies. Analysts were left in the dark on such basic issues in the past, which helps explain why bogus information from fabricators was included in some prewar intelligence reports on Iraq. And, when they write their reports, they must include better attribution and sourcing for each major assertion.

“I think the Iraq experience gave them thicker skins,” said one former senior intelligence official, who like several others quoted in this article would speak only on the condition of anonymity about internal agency matters. “There was a lot of work done to tighten up the tradecraft.”

Unlike the prelude to the Iraq war, when many critics accused Bush White House officials of cherry-picking the intelligence to conform to their policy, some outside analysts say they do not see evidence of the Obama administration pushing intelligence officials to come up with predetermined answers.

“The intelligence was so heavily politicized on Iraq,” said Joseph Cirincione, the president of the Ploughshares Fund, a global security foundation. “The higher up the chain in the government the intelligence reporting went, the more it got massaged, and the doubts and caveats got removed.”

But now, he said, “I haven’t heard any complaints about the administration pressuring the intelligence community to tilt the intelligence.”

He added that while conservative political leaders in the United States and Israel had complained about the intelligence assessments on Iran, such outside criticism did not have the same impact it would coming from the White House.

“It’s one thing to have the prime minister of another country come to town and say time is short, but it’s another thing to have the vice president go to Langley and pressure people,” he said, referring to former Vice President Dick Cheney’s repeated visits to CIA headquarters before the Iraq war.

But some conservatives who support more aggressive action to prevent Iran from gaining a nuclear weapon argue that the CIA’s restraint has, in fact, been influenced by political pressure exerted by the Obama administration. President Barack Obama has said he would use military force only as a last resort against Iran, and conservatives argue that the Obama administration does not want the intelligence community to produce any reports suggesting the Iranians are moving swiftly to build a bomb.

“The intelligence analysts I’ve dealt with have always been willing to engage in debates on their conclusions, but there is top-down pressure to make the assessments come out a certain way,” said John R. Bolton, a senior fellow at the American Enterprise Institute and a former ambassador to the United Nations in the Bush administration.

Memories of Iraq have clouded the debate on Iran ever since the United States intelligence community first concluded in 2007 that Iran had halted its weapons program four years earlier. In late 2007, Michael McConnell, then the director of national intelligence, took a new National Intelligence Estimate — the consensus of analysts at the government’s 16 intelligence agencies — to the White House to brief President George W. Bush about the report’s startling new findings.

Officials at the White House, still stung from the criticism on Iraq, quickly realized that they would face a firestorm of protest if they did not make the findings public, according to former administration and intelligence officials. A classified version of the new assessment would go to the congressional intelligence committees, where lawmakers would see that analysts had reached a sharply different conclusion about Iran than they had two years earlier, when they had concluded that Iran’s weapons program was still under way.

News of the shift would probably leak to the news media, and White House officials feared that the administration would then be accused of suppressing intelligence on Iran, just as it had been criticized for doing on Iraq, according to the former officials. White House officials also worried that they would be accused of tainting the intelligence process, so they pressed McConnell to have the intelligence community write and issue its own declassified summary.

Some senior intelligence officials who rushed to write the document over a weekend objected to disclosing their conclusions, but to no avail. “I was told that I didn’t get it; this wasn’t a request,” one official recalled.

Once published, the report created an uproar. Conservative critics blasted CIA officials, saying that the intelligence community was freelancing and trying to influence the political debate, and to make up for its shortcomings on Iraq by now trying to stop a war with Iran. Among them was Bolton, who dismissed the 2007 assessment as “famously distorted” and called on Congress to investigate its politicization.

Stung by those attacks, and the aftershocks of the poisonous political debate over the role of the intelligence community on Iraq and Iran, officials at the CIA and other agencies did not release a public version of the 2010 assessment on the Iranian nuclear program, which concluded that while Iran had conducted some basic weapons-related research, it was not believed to have restarted the actual weapons program halted in 2003.

Thomas Fingar, who was chairman of the National Intelligence Council at the time of the 2007 assessment on Iran, said that analysts had to be willing to make tough calls based on fragmentary evidence, and not get distracted by what he called the rare instances of political pressure or their own previous lapses.

“Learning from past mistakes is imperative,” he said. “Worrying about them is pointless.”

© 2012 The New York Times Company