Did torture "work" on Abu Zubaydah?

updated 4/22

Did torturing Abu Zubaydah yield valuable information? Former CIA Director Michael Hayden and former Attorney General Michael Mukasey, writing in the Wall Street Journal, say yes (April 17):

The terrorist Abu Zubaydah (sometimes derided as a low-level operative of questionable reliability, but who was in fact close to KSM and other senior al Qaeda leaders) disclosed some information voluntarily. But he was coerced into disclosing information that led to the capture of Ramzi bin al Shibh, another of the planners of Sept. 11, who in turn disclosed information which -- when combined with what was learned from Abu Zubaydah -- helped lead to the capture of KSM and other senior terrorists, and the disruption of follow-on plots aimed at both Europe and the U.S. Details of these successes, and the methods used to obtain them, were disclosed repeatedly in more than 30 congressional briefings and hearings beginning in 2002, and open to all members of the Intelligence Committees of both Houses of Congress beginning in September 2006. Any protestation of ignorance of those details, particularly by members of those committees, is pretense.

An unnamed "former intelligence official" tells Scott Shane of The New York Times no:
(April 18):

Abu Zubaydah had provided much valuable information under less severe treatment, and the harsher handling produced no breakthroughs, according to one former intelligence official with direct knowledge of the case. Instead, watching his torment caused great distress to his captors, the official said.

The Times account continues:

Through the summer of 2002, Abu Zubaydah continued to provide valuable information. Interrogators began to surmise that he was not a leader, but rather a helpful training camp personnel clerk who would arrange false documents and travel for jihadists, including Qaeda members.

He knew enough to give interrogators “a road map of Al Qaeda operatives,” an agency officer said. He also repeated talk he had heard about possible plots or targets in the United States, though when F.B.I. agents followed up, most of it turned out to be idle discussion or preliminary brainstorming.

At the time, former C.I.A. officials say, his tips were extremely useful, helping to track several other important terrorists, including Mr. Mohammed.

Ambiguity as to the possible role of torture persists here. Zubaydeh was captured on March 28, 2002. The memo authorizing the full package of techniques used on Zubaydeh is dated Aug. 1, 2002. The Times claims that he revealed information "throughout the summer." It also reports that after Zubaydah named KSM as the main organizer of the 9/11 plot to FBI agents,

A C.I.A. interrogation team that arrived a week or two later, which included former military psychologists, did not change the approach to questioning, but began to keep him awake night and day with blasting rock music, have his clothes removed and keep his cell cold.

It's agreed that Zubaydah provided information that helped to capture Khalid Shaikh Mohammed. And there's no dispute that he was eventually tortured. But when did techniques brought into his interrogation-in-progress by the CIA reach the level of torture, and what did Zubaydah suffer prior to revealing the information that led to KSM's capture?

It's important not to conflate Zubaydah's revealing KSM's role in the 9/11 attacks, which according to the Times happened before the CIA interrogators arrived, with the information he provided that "helped to track" other terrorists, including KSM (according to Mukasey and Hayden, by providing info leading to the capture of Ramzi bin al Shibh).

In short, these two apparently diametrically opposed accounts do not flat-out contradict each other.

4/22 UPDATE: Jane Mayer, citing the just-released Levin Report, adds this to the chronology:

By June 2002—again, months before the Department of Justice gave the legal green light for interrogations—an F.B.I. special agent on the scene of the interrogation of Abu Zubaydah refused to participate in what he called “borderline torture,” according to a D.O.J. investigation cited in the Levin report. Soon after, F.B.I. Director Robert Mueller commanded his personnel to stay away from the C.I.A.’s coercive interrogations.

Still unclear: did information gain by torture or "borderline torture" lead to the capture of Ramzi bin al Shibh and, by extension, Khalid Shaikh Mohammed?

It would be deeply gratifying to establish definitively that the torture of detainees yielded no valuable information. But Americans' rejection and prosecution of these acts may well have to integrate at least equivocal evidence that coercive interrogation sometimes yields actionable intelligence. We will have to accept that any information gained through torture is not worth the price -- that investing our government and military with complete power over the bodies of people held in custody is a greater threat to our society than terrorist attack.

Continued:
Did torture "work" is the wrong question - Ali Soufan shows why
Depends what your definition of "works" is