The “Dirty 30” probably weren’t all Osama bin Laden bodyguards after all. The “Karachi 6” weren’t a cell of bombers plotting attacks in Pakistan for al-Qaida. An Afghan man captured 14 years ago as a suspected chemical weapons maker was confused for somebody else.
An ongoing review shows the U.S. intelligence community has been debunking long-held myths about some of the “worst of the worst” at Guantánamo, some of them still held today. The retreat emerges in a series of unclassified prisoner profiles released by the Pentagon in recent years, snapshots of much larger dossiers the public cannot see, prepared for the Periodic Review Board examining the Pentagon’s “forever prisoner” population.
Afghan Abdul Zahir is a case in point: U.S. Rangers captured him and some “suspicious items” in a July 11, 2002, raid on his home on suspicion of “involvement with chemical/biological weapons activity.” During George W. Bush’s presidency he was briefly charged with war crimes, accused of being an al-Qaida conspirator named Abdul Bari, a nickname used by Zahir as well. The parole board cleared him for release on July 11, 2016 — no trial necessary — after an intelligence assessment concluded he “was probably misidentified as the individual who had ties to al-Qaida weapons facilitation.”
“They had the wrong guy the whole time,” said Air Force Lt. Col. Sterling Thomas, his defense attorney since 2010. “Abdul Zahir shared a name with a terrorist that they thought they were looking for. He unfortunately was further condemned by the fact that United States forces couldn’t distinguish between bomb-making materials and the salt, sugar and petroleum jelly he had nearby when he was wrongly arrested.”
Sign Up and Save
Get six months of free digital access to the Miami Herald
The new intelligence reports are not designed to help the panel decide a captive’s guilt or innocence. Rather they were prepared for representatives from the Departments of Defense, Justice, Homeland Security, State and the Director of National Intelligence to evaluate each captive, a process that has whittled the detainee population down to 61 today.
They had the wrong guy the whole time. Abdul Zahir shared a name with a terrorist that they thought they were looking for.
Air Force Lt. Col. Sterling Thomas, defense attorney
In the instance of Zahir, the parole panel said the 44-year-old Afghan had a “limited role in Taliban structure and activities” and approved his release with “appropriate security assurances.” The State Department is pursuing a plan to resettle or repatriate him.
“The National Counterterrorism Center compiles the detainee compendium consisting of a presentation of specific facts that includes information about the detainee and, where appropriate, notes inconsistent reporting,” said Army Lt. Col. Valerie Henderson, the Pentagon’s public affairs officer responsible for Guantánamo policy.
The documents also offer a window into the wobbly world of early war-on-terror intelligence gathering and analysis where a suspicion built on circumstances of capture gelled into allegations of membership in a terror cell that on reflection more than a decade later probably didn’t exist. In a series of interviews, intelligence sources — including people who served at Guantánamo at the time — blamed bad intelligence on a combination of urgency to produce, ignorance about al-Qaida and Afghanistan at the prison’s inception and inexperience in the art of investigation and analysis.
“It was clear early on that the intelligence was grossly wrong,” said Mark Fallon, a retired 30-year federal officer who between 2002 and 2004 was Special Agent in Charge of the Department of Defense’s Criminal Investigation Task Force. Most “weren’t battlefield captives,” he said, calling many “bounty babies” — men captured by Afghan warlords or Pakistani security forces and sent to Guantánamo “on the sketchiest bit of intelligence with nothing to corroborate.”
Once they got to Guantánamo, he said, the task of discerning tidbit from truth was complicated by an Army leadership unskilled in military intelligence arguing that “you couldn’t coddle the enemy.” It was an era of since-documented manipulation of temperatures, sleep, diet and solitary confinement to get a captive talking at a time when Fallon’s team was advocating rapport building. Sleep deprivation only muddles memories, he said. So they ended up with “a lot of false information based on some pretty poor interrogations being done partly by military interrogators in that time frame.”
It was clear early on that the intelligence was grossly wrong.
Mark Fallon, Special Agent in Charge of the Department of Defense’s Criminal Intelligence Task Force, 2002-04
Fallon was responsible for some interrogations and evaluating intelligence with an eye toward prosecution by military commission. Now, more than decade later, he is in the final stages of publishing a book of his criticisms and said in a recent interview that it’s no surprise that early prisoner profiles are imploding under Periodic Review Board scrutiny.
“That’s why people are so successful doing cold case homicide cases,” he said. “People make decisions based on what they knew then. I don’t want to say that the facts changed. The facts grew. When you’re working cases, cases evolve. As you get additional facts, you interpret it differently.”
The current director of intelligence at the prison, Navy Capt. Kathleen Hawk, declined repeated requests for an interview to discuss the state of intelligence gathering in light of the retreats from earlier profiles.
In the early years, according to one analyst who worked there, Guantánamo’s Joint Intelligence Group was “looking for anything you can pin on these guys.” He still works in the U.S. intelligence community, and spoke to the Herald at length on condition he not be named.
Analysts “weren’t making things up,” said the soldier-turned-civilian contractor who worked at the JIG for four years. “We were working overtime trying to figure out who we had, learning this culture as we go along.” The blended military-contractor unit was isolated, started off knowing more about Russia and Bosnia than Afghanistan and al-Qaida and was under pressure to help stop the next terror attack.
“Everybody’s drinking the Kool-Aid. You see conspiracies everywhere,” said the analyst. The intelligence unit was “picking up on one or two things and holding on to it tightly like it was gospel.”
That’s how it happened that, at Guantánamo, being captured with a cheap Casio watch on your wrist made you a terrorist.
It worked this way: Analysts were told that an al-Qaida bomb-making course gave its graduates Casio F-91W or A159W models as gifts. Separately, it was understood that those models could be used as detonation devices. So possession of one became “an indicator of al-Qaida training in the manufacture of improvised explosive devices,” according to a 2006 intelligence unit Matrix of Threat Indicators for Enemy Combatants provided to the Herald in 2011 by WikiLeaks.
“Absolutely ridiculous,” said the analyst, calling it “an awfully big leap.”
Especially since a footnote in the very same matrix explained that “approximately one-third” of Guantánamo prisoners who were captured with that ubiquitous model of watch “have known connections to explosives.” Meaning two-thirds did not.
“They would take any tidbit of information and use that. It’s what the psychologists call confirmation bias,” said Fallon of the early flawed intelligence. “If you are predisposed that that person never leaves Guantánamo, you make judgments and conclusions on those facts based on that predisposition. We were looking for corroborated information, not just a tidbit. That’s why you do investigations, not just interrogations, to corroborate that information.”
There was also pressure to defend the use of Guantánamo. The “rush to justify the place as gathering ‘actionable intelligence’ ” — an expression that 2002-04 prison commander Army Maj. Gen. Geoffrey Miller favored in talking to reporters — did not “allow for more sober thoughtful analysis,” said Fallon.
Fear of recidivism also infected the process. Some in the intelligence unit were reluctant to discredit earlier information, which could lead to an angry captive’s release. “If this guy wasn’t an enemy before, he is now,” said the former Guantánamo analyst, describing a genuine concern that clouded whether to question a prisoner’s intelligence profile.
It was prescient, if beyond the scope of their jobs. A recent report from the National Director of Intelligence showed that, as of July 15, the U.S. intelligence community has concluded that 17.6 percent of former captives took part in terrorist or insurgent activities after their release. Most were let go from Guantánamo during the Bush years. It said that 30 of the confirmed 122 recidivists are dead and 25 were subsequently captured by presumably allied countries, meaning 67 recidivists were at large.
The intelligence community found another 12.4 percent, or 86 former captives, were suspected of “returning to the battlefield,” as critics of release call it — 21 of them captured or dead, and 65 of them at large.
We started looking at the detainees and we realized these guys were not all the same. There were important differences among the detainees, the role they played and threat they posed.
Matt Olsen, director of an interagency task force in 2009
Even by 2009, when an interagency task force dug through the remaining 240 detainee files, “we struggled at times to definitively identify detainees because of the multiple names they used. There were a few pretty complicated cases,” said then task force Director Matt Olsen, who went on to run the National Counterterrorism Center that today prepares prisoner profiles for the parole-style board.
The 2009 review reached beyond Defense Department analysis to include material from the CIA and other agencies, he said. There were labels, like “worst of the worst,” Olsen said, some of them “myths” that would be debunked. “We started looking at the detainees and we realized these guys were not all the same. There were important differences among the detainees, the role they played and threat they posed.”
“We had to dig into terms like ‘associated with al-Qaida,’ for example, to understand exactly what that meant from a threat perspective and whether those links were strong or tenuous.”
One file that may have eluded 2009 understanding belonged to Mustafa al Shamiri. His leaked classified 2008 assessment cast the Yemeni as a “high risk” captive of consequence, a senior al-Qaida training camp trainer and guesthouse logistician. Olsen’s team declared him too dangerous to release — one of 48 indefinite detainees in the war on terror, a forever prisoner of the forever war.
But the Periodic Review Board on Jan. 12 declared earlier intelligence “discredited.” Earlier intelligence analysts had confused him for someone of a similar name. Cleared for release, he is still at Guantánamo.
The findings beg the question of who got it wrong, and when. Was it then, when the U.S. military was managing the intelligence assessments at Guantánamo? Or is it now with the National Counterterrorism Center combing through intelligence to build detainee dossiers?
Cully Stimson, the Pentagon’s deputy assistant secretary of defense for detainee affairs in 2006 and 2007, recalls that the intelligence was at times puzzling enough that he’d go to Guantánamo to study material in the JIG’s evidence locker. Shelves of the prefab building in the Detention Center Zone were stacked with material that arrived on the cargo planes carrying the captives from Afghanistan — from wallets and wristwatches to notebooks and family photos.
His job included recommending to then-Deputy Defense Secretary Gordon England whether to release certain captives. A former prosecutor and defense lawyer, Stimson said he was delving deeper than written reports to try to distinguish tip from truth in a captive’s file.
“The skeptical side of me says the scales are tipped in the favor of transfer and they’re going to do what they need to do bureaucratically to get it done,” says Stimson, discussing release decisions based on updated intelligence. “I would hate to think that people are shading or suppressing raw intelligence to the contrary. At the same time what I don’t know is if additional information has been developed over time that would lead a reasonable person to that conclusion. If that’s the case, then it is what it is.”
The former intelligence analyst said it’s not that the Guantánamo intelligence shop didn’t have an inkling of errors in the early years. He recalled analysts writing assessments in a secure facility poking fun at one another. “How’s your kingpin over there?” But it was hard to be the spoiler who poked holes in a previous analyst’s pronouncements.
Plus, according to Fallon, troops in Afghanistan sent captives to Cuba with “no effective screening mechanisms,” meaning analysts had to start from scratch trying to figure out who the U.S. held in the cages at Camp X-Ray. “Guantánamo became more of a dumping ground rather than a place, at least from my position, what we were told was supposed to be a facility for high-value targets for either intelligence or prosecution.”
That’s how 30 early captives became known as “The Dirty 30” and a different half-dozen captives were called the “Karachi 6.”
The intelligence unit was sorting files, and somebody tasked with finding bodyguards noticed the common capture of 30 men delivered by Pakistani security forces and so labeled them. But by the time the files leaked, in 2011, a review of prison and court records shows, the Bush administration had repatriated 10 of them, including a Moroccan profiled as a leader of the Dirty 30 in July 2003. Updated, unclassified profiles cast doubt on whether some still were bodyguards — or that aspect of the assessment has simply vanished.
The leaked 2006 “Matrix of Threat Indicators” — basically a guide to assessing a detainee’s dangerousness — took as an act of faith the existence of a unit called the “Karachi 6,” six men captured on Sept. 11, 2002, and profiled as a would-be terrorist bombing cell.
Then in December 2015, the intelligence community debunked the idea of the Karachi 6 in a series of intelligence profiles, including this one about Yemeni Ayub Salih.
“We judge this label more accurately reflects the common circumstances of their arrest and that it is more likely the six Yemenis were elements of a large pool of Yemeni fighters that senior al-Qaida planners considered potentially available to support future operations. Our review of available intelligence indicates that he probably did not play a major role in terrorist operations, leading us to disagree with previous U.S. government assessments that he was involved in a 2002 plot to conduct an attack in Karachi, Pakistan.”
The board process has also disclosed new information that may have eluded the prison’s intelligence unit. Even so that hasn’t always led to a captive’s release.
That’s what happened in the instance of Saudi Mohammed al Qahtani, held as a forever prisoner and suspected would-be 20th hijacker in the Sept. 11, 2001, terror attacks. The board declared him too dangerous to release but acknowledged something that his leaked 2008 prison profile declaring the “detainee is in good health” didn’t know, or didn’t include:
The man who was subjected to such cruel interrogation at Guantánamo that he was disqualified from trial as an alleged 9/11 plotter had a history of profound mental illness. His lawyers provided the panel with an affidavit from an American psychiatrist who treats U.S. war veterans saying that Qahtani had been sick since his childhood in Saudi Arabia. He suffered schizophrenia, major depression, a traumatic brain injury and had been hospitalized for it, according to the doctor, Emily Keram.
Oops — examples from the intelligence files
▪ Yemeni Moath al Alwi got to Guantánamo the week Camp X-Ray opened and was profiled seven years later as a veteran jihadist, combatant in the battle for Tora Bora and former Osama bin Laden bodyguard — a member of the “Dirty 30.” A July 2015 intelligence profile prepared for that file review said that, before his capture by Pakistani forces in December 2001, Alwi was “probably not” one of bin Laden’s bodyguards but concluded he was “an al-Qaida-affiliated fighter who spent time with” bin Laden’s security detail.
▪ Yemeni Shawki Balzuhair got to Guantánamo in 2002 and was held as a member of the Karachi 6. He “was probably awaiting a chance to return to Yemen when he was arrested” at a Karachi safehouse on Sept. 11, 2002, a new intelligence assessment wrote in January. He was cleared and remains at Guantánamo.
▪ Yemeni Mansoor al Dayfi got to Guantánamo less than a month after Camp X-Ray opened, and a 2008 profile called him an al-Qaida commander who knew about the 9/11 attacks before they happened. His July 2015 profile, however, described him as probably “a low-level fighter” and questioned whether he was a member of al-Qaida at all, and made no mention of foreknowledge of the Sept. 11 attacks. It concluded “he probably has exaggerated his involvement in and knowledge of terrorist activities during some of his interrogations.” He was released to Serbia in July.
▪ Yemeni Saeed Jarabh got to Guantánamo in February 2002 and was profiled at the prison in 2008 as a high-risk former bin Laden bodyguard. His 2015 profile made no mention of it, unlike other once-profiled bodyguards, and approved his release calling him “a low-level fighter” who “lacked a leadership position in al-Qaida or the Taliban.” He was cleared in March 2015 and sent to a rehabilitation program in the United Arab Emirates in October.
▪ Yemeni Hayl Maythali, another Karachi 6 captive, held at Guantánamo since October 2002, “probably acted briefly as a guard” at a bin Laden compound in Kandahar, but a March 7, 2006, reassessment retreated from Karachi terror cell membership. It said he “was probably awaiting a chance to return to Yemen when he was arrested” at a Karachi safe house, rather than being “part of an al-Qaida operational cell intended to support a future attack.” He was cleared and is at Guantánamo.
▪ Yemeni Said Nashir, whose lawyers call him Hani Saleh Rashid Abdullah, was held at Guantánamo as a member of the Karachi 6. A December 2015 assessment recast him as “probably intended by al-Qaida senior leaders to return to Yemen to support eventual attacks in Saudi Arabia.” But, it noted, he “may not have been witting of these plans.” The board has yet to decide whether to declare him too dangerous to release.
▪ Kuwaiti Fawzi al Odah got to Guantánamo in February 2002 and was profiled at the prison in 2008 as a bin Laden “associate” and one-time London-based “extremist recruiter and courier.” But a March 2014 intelligence assessment declared, “We lack confidence in statements from other detainees that [he] was closely associated with” bin Laden “or belonged to an al-Qaida cell in London.” The parole board approved his release four months later, noting the Kuwaiti’s “low level of training and lack of leadership in al-Qaida or the Taliban.” The board cleared him four months later, and he was repatriated the same year.
▪ Yemeni Mashur al Sabri got to Guantánamo in May 2002 and was profiled in 2008 as “a member of a Yemeni al-Qaida cell directly involved with the USS Cole attack.” A new assessment in 2014 said he “had ties to numerous extremists but probably did not play a significant role in terrorist operations.” As for al-Qaida’s Oct. 12, 2000, bombing of the warship off Aden, Yemen, that killed 17 sailors, “there is no indication [he] had foreknowledge of the attack.” He was sent to a rehabilitation program in Saudi Arabia in April.
▪ Yemeni Ayub Ali Salih got to Guantánamo in 2002 and was profiled as a member of the Karachi 6 terror cell, supposedly plotting attacks in Pakistan at his capture on Sept. 11, 2002. A reassessment dated December 2015 declared that based on “our review of available intelligence,” he “probably did not play a major role in terrorist operations, leading us to disagree with previous U.S. government assessment that he was involved in a 2002 plot to conduct attacks in Karachi, Pakistan.” He was cleared within months and sent to a rehabilitation program in the United Arab Emirates in October.
▪ Yemeni Mustafa al Shamiri got to Guantánamo in June 2002, and his 2008 prison profile cast him as a captive of consequence, a senior al-Qaida trainer and guesthouse logistician. When the board cleared him for release eight years later, it disclosed that earlier “derogatory prior assessment” of him had been “discredited” and he was little more than a “low-level fighter.” Whoever assembled his profile confused him for someone of a similar name. He’s still at Guantánamo.
▪ Afghan Abdul Zahir got to Guantánamo after his July 11, 2002, capture in a raid that targeted a man named “Abdul Bari,” who was “believed to be involved in the production and distribution of chemical or biological weapons for al-Qaida.” A 2015 profile did not discredit his ties to al-Qaida entirely but said the now 44-year-old Afghan was “probably misidentified as the individual who had ties to al-Qaida weapons facilitation activities.” One data point: At his capture, “U.S. forces recovered samples of unknown substances in the raid, including a white powder, that were initially believed to be chemical or biological agents, although other information later proved the samples to have been salt, sugar, and petroleum jelly.”