The book also reports that Al-Qaeda's second-in-command, Ayman al-Zawahri, canceled the plan just 45 days before it was to have been carried out. U.S. government officials have not commented on the account in the book, titled "The One Percent Doctrine," but the New York Police Department has confirmed that it knew of the plot.
Data from a confiscated computer that revealed the planned attack to release hydrogen cyanide in the New York subway system, according to the book, excerpts of which were published this week in the weekly American news magazine, "Time."
Author Ron Suskind writes that the data on the computer revealed that Al-Qaeda had developed a canister that could be remotely activated to release the deadly poison.
Potentially As Bad As 9/11
The book says U.S. officials believed such an attack could have a death toll of 3,000 -- including 1,000 trampled to death by panicked commuters. That would be close to the number of people killed in the attacks on New York and Washington on September 11, 2001.
More frightening still, the book says, is that al-Zawahri eventually called off the attack. U.S. President George W. Bush and Vice President Dick Cheney discussed the failed plot at several White House meetings with intelligence officials. At one meeting, Suskind writes, Bush worried that it might mean Al-Qaeda was planning an even bigger strike.
Word of the cancellation came from an informant identified only as Ali, who is said to be close to Al-Qaeda leaders. The book says Bush was also worried about why a devoted militant would help America. That question remains unanswered.
Old-Fashioned Cloak-And-Dagger Spying
After the September 11, 2001, attacks, U.S. intelligence agencies were criticized for failing to anticipate the terrorist operation. Some said that after the Cold War, the intelligence community relied too much on high technology -- satellites and electronic eavesdropping -- at the expense of old-fashioned cloak-and-dagger spying.
However, these same critics concede that getting "human intelligence" on Al-Qaeda is virtually impossible because such groups recruit only those who are close friends or family members.
So how did U.S. intelligence manage to get human intelligence about Al-Qaeda? Suskind writes that there was, in his words, "disgruntlement" within Al-Qaeda after U.S.-led forces had routed it and the Taliban from power in Afghanistan.
And "disgruntlement" is the key, according to Simone Serfaty, a foreign-affairs analyst who specializes in intelligence matters at the Center for Strategic and International Studies, a policy research center in Washington.
"The question is the extent to which this is just an isolated case which fell into our laps or whether it is something that we worked on and [something that] proved to be a point of departure for many more such activities over the past three years," Serfaty said.
Retired U.S. Army General Edward Atkeson, who served as a senior military-intelligence officer during the Cold War, says the United States has a history of offering very attractive inducements to disgruntled enemies.
Atkeson said the informant in the New York subway plot was Pakistani, and therefore probably educated at one of that country's many strict madrasas, or Islamic schools.
"But that doesn't mean that the guy doesn't continue to have thoughts passing through his mind as to which is more gainful or beneficial for him personally: to follow that path [of Al-Qaeda] and become a fighter, or to turn against it and take advantage of whatever the Americans have to offer," Atkeson said.
Atkeson points to the generous rewards that were offered for the capture of deposed Iraqi President Saddam Hussein and Abu Mus'ab al-Zarqawi, the late head of Al-Qaeda in Iraq.
Both Atkeson and Serfaty agree that the intelligence on the planned subway attack seemed to be good -- unlike the intelligence about Hussein's suspected weapons program that prompted Bush to order the invasion of Iraq in March 2003.
Importance Of Good Analysis
Intelligence is more an art than a science, Atkeson notes. Certainly many scientific methods are used in the gathering of intelligence, but he says it is how analysts judge this data that matters.
"Analysis is part of the intelligence process," he said. "Intelligence is not just the collection of information. The intelligence process is the assessment of it all and the winnowing down and beating it against evidence of a different nature."
Serfaty says such judgments can be very, very wrong. He points not only to the faulty prewar intelligence on Iraq, but to many well-known failures during the Cold War, including a U.S. conclusion -- inaccurate, as it turned out -- that the Soviet Union wouldn't put missiles in Cuba.
Serfaty says all intelligence is, by its nature, incomplete. But he emphasizes that those who gather these fragments of facts aren't to blame for bad decisions by a country's leaders.
"I think it's not a failure of intelligence that we have faced over the past several years," Serfaty said. "It's a failure of analysis. And why? Because a lot of those people were not competent as analysts. And why were they not competent? Either because they did not have adequate training or because they had prejudices. That is to say they started with a conclusion and on that basis could not do adequate analysis."
But no matter how accurate the intelligence was on the planned attack on the New York subways, good spying didn't stop it. The plot was canceled by al-Zawahri himself, for reasons that are still unknown.