Consider, for example, the findings of an internal CIA investigation in 1995. After the Soviet Union collapsed, the CIA's inspector general examined how in the late 1980s and early 1990s the CIA had incorporated Russian disinformation into its own reporting. He discovered that over those years the KGB had dispatched at least a half-dozen double agents who provided disinformation cooked up in Moscow to their CIA case officers. Between 1986 and 1994, some of this data had routinely been passed to Presidents Ronald Reagan, George H.W. Bush and Bill Clinton in reports with a distinctive blue stripe to signify their importance.
When the inspector general traced the path of this disinformation, he found that the "senior CIA officers responsible for these reports had known that some of their sources were controlled by Russian intelligence." CIA Director John Deutch, who had received the blue-border reports when he was deputy secretary of defense, told Congress that the CIA's failure to disclose that the intelligence emanated from KGB-controlled agents was "an inexcusable lapse."
The only way that the KGB could have duped the CIA for years was by modifying its data so that it would continue to seem plausible—and that required some form of feedback. As it later turned out, the KGB had no fewer than three moles in American intelligence capable of providing such feedback: In the CIA it had Aldrich Ames starting in 1985. And in the FBI the KGB had both Robert Hanssen since 1978 and Earl Edwin Pitts starting in 1987. They survived as moles—Hanssen for 22 years—because of the sort of institutional blindness, born out of bureaucratic fear, so well described in Le Carre's novel. Those officers in the CIA who attempted to remedy this blindness, notably counterintelligence chief, James Jesus Angleton, were dismissed as paranoid.
These double agents came to light largely because of the defections from the KGB that followed the fall of the Soviet Union. Unfortunately, under more normal circumstances, entrenched bureaucracies can be expected to resist reappraisals of their past work, especially where careers are at stake. The intelligence community's 2007 National Intelligence Estimate (NIE) on Iran is a case in point.
Based on intelligence, including reports from agents and defectors, that an Iranian nuclear weapon-design program—code-named Project 111—had ended, the NIE declared: "We judge with high confidence that in fall 2003, Tehran halted its nuclear weapons program," including "nuclear weapon design and weaponization work and covert uranium enrichment." The intelligence community took at least partial credit for this success by attributing Iran's change to "increasing international scrutiny and pressure resulting from exposure of Iran's previously undeclared nuclear work."
Today no one, including the International Atomic Energy Agency, believes that Iran gave up its nuclear weaponization ambitions. Indeed we now know from satellite imagery and other means that in 2003 the regime was secretly completing a new uranium-enrichment facility at Fordo, 20 miles north of the holy city of Qom. That was after it closed down Project 111, which in any case had been compromised by a laptop stolen from Iran and smuggled into Turkey and then into CIA hands.
Nor can the CIA rely on its own espionage apparatus, because a communications accident in 2004 compromised most, if not all, of its agents in Iran: The CIA inadvertently sent a list of its operatives to a double agent, a disaster described by the reporter James Risen in his book "State of War." As a result, the CIA could not be sure how much of the data it received from those operatives was disinformation.
Yet, as far as is known, the CIA has still never reappraised the sources and methods that led to its conclusion that Iran had abandoned its quest for a nuclear weapon.
Mr. Epstein's latest book is "James Jesus Angleton: Was He Right?"
