Pakistani Bounty Claims: Adnan Farhan Abd Al Latif and TD-314/00684-02

Benjamin Wittes has been complaining that no one besides Lawfare’s writers is looking closely at the DC Circuit decision in Latif.

Why has there been virtually no press coverage of the Latif decision? Other than this article on CNN’s web page, which actually ignores the aspect of the case that makes it jurisprudentially important, a search on Google News reveals none (other than Lawfare stuff).

Memo to the press: This case is important. It is far more likely, in my judgment, to provoke a cert grant than any habeas case the D.C. Circuit has decided to date. If and when it does so, it will present a novel and deeply important question to the Supreme Court: Whether the courts in reviewing these habeas cases should start with a presumption of the validity of government intelligence.

So I decided to take a closer look.

At issue is a Yemeni detainee, Adnan Farhan Abd Al Latif, picked up in Pakistan in December 2001 and alleged to have trained with al Qaeda. Judge Henry Kennedy granted Latif’s habeas petition last summer, largely because he found the government’s single most important piece of evidence–an intelligence report of some kind (which I’ll call the Report)–unreliable. The DC Circuit–with Judges Janice Rogers Brown and Karen Henderson in the majority and David Tatel in dissent–remanded the case on the issue of the detainee’s credibility. But on the more important issue–whether Kennedy was correct in dismissing the Report–they overturned the district decision. Here’s Wittes’ description of the evidentiary issue.

I think the document in question is a report with the serial number TD-314/00684-02 that I take to be the CIA’s report of Pakistani claims about a significant number of detainees they turned over to the US in December 2001–basically the intake report for a chunk of detainees, possibly (given the time and place) turned over for bounty.

Here’s my logic: Latif’s Gitmo file makes the same claim the government made in his habeas case: that Latif trained and then fought with al-Qaeda. But that entire report cites just one source–TD-314/00684-02–to make that claim. It cites TD-314/00684-02 to support the following assertions:

While detainee was with the Taliban, he encountered Abu Hudayfa the Kuwaiti; Abu Hafs the Saudi, and Abu Bakr from the United Arab Emirates or Bahrain. Detainee claimed he saw a lot of people killed during the bombings, but never fired a shot. Detainee then traveled to Jalalabad, AF, and crossed into Pakistan with fleeing Arabs, guided by Taqi Allah.


Detainee’s recruiter is assessed to be senior al-Qaida facilitator Ibrahim Muhammad Abd al-Razzaq Baalawi, aka (Abu Khulud). Detainee admitted Ibrahim Aliwee convinced detainee to travel to Afghanistan for jihad and admitted staying at Abu Khulud’s residence for a short period in Kandahar.


Detainee admitted receiving weapons training from the Taliban and then fighting in support of the Taliban on the front lines.


Detainee admitted after training he was sent to the front lines north of Kabul. Detainee remained there until the Taliban retreated and Kabul fell to the Northern Alliance.

For the remaining assertions regarding Latif’s ties to Al Qaeda, the Gitmo report includes no citation.

Given that Latif has consistently denied fighting with al Qaeda since he’s been in US custody, instead telling a reasonably consistent story about traveling to Afghanistan for medical care for a head wound, it raises questions about the circumstances under which “he admitted” to fighting with the Taliban. And there’s apparently just this one report supporting that claim. Moreover, if the government has such evidentiary problems in this case, you would think they’d cite this report if they hadn’t already. So I assume they have, and that’s the report that Kennedy found to be so problematic.

I surmise that the report was the intake report for a bunch of detainees turned over by the Pakistanis by looking at the eight other Gitmo reports that cited this document (note, for the most part, I use the name that Andy Worthington did in his invaluable definitive list of Gitmo prisoners):

The citations in these detainees’ Gitmo files relying on TD-314/00684-02 pertain to whether he was involved in the fighting and how he came to be turned into the Pakistanis. And all of the detainees (save Sultan, in which the citation to TD-314/00684-02 pertains to a different detainee) were captured at the Pakistani border in December 2001 and turned over to the Americans shortly thereafter. I’m not sure, but I think all of them may have been captured in Kohat. Assuming that the report’s low serial number (00684) means it precedes a slightly higher one written around January 5, 2002, this report appears to have been written in the first 5 days of 2002.

So if I’m right, then the report the DC Circuit is fighting over is the CIA intake report for a bunch of detainees captured in December 2001 and turned over at the end of that month.

There are a number of reasons why such a report, by itself, would be unreliable. First, the sheer number of men being transferred from Pakistani to US custody at the time–over a hundred in a bunch. To make things worse, in the process of transferring these detainees to Peshawar, a busload tried to escape, killing 6 Pakistani guards and a number of detainees (some of the detainees whose reports cite TD-314/00684-02 are alleged to have been among those who tried to escape). In addition, there’s the likelihood the Pakistanis did the interrogations. Since Pakistanis were getting bounties for each “al Qaeda fighter” they turned over, they had an incentive to claim the detainees had been fighters. Even worse, for the Arabic men in question, the interrogation would have required translation between Arabic and Urdu or Pashto (for the interrogation itself), and then translation from that into English (to go into the CIA report). There are all sorts of reason why you shouldn’t indefinitely detain a man based on such a report!

In any case, there are several things Tatel said (at times citing portions of Kennedy’s ruling that were redacted in the original) that suggest this report, or something just like it, is the report at issue.

Tatel hints at the condition of production of the report.

By contrast, the Report at issue here was produced in the fog of war by a clandestine method that we know almost nothing about.


[The] intelligence report, was, in this court’s own words, “prepared in stressful and chaotic conditions, filtered through interpreters, subject to transcription errors, and heavily redacted for national security purposes.”

As Tatel describes, the documents “contain multiple levels of hearsay, depend on translators of unknown quality, and include cautionary disclaimers”

And then Tatel cites a number of precedents that would be pertinent to such a report: he describes Barhoumi, in which the government didn’t have (or turn over) the original Arabic version of a diary; he describes Bensayah and Al Alwi, in which the court scrutinized whether interrogation reports of a detainee’s statements were accurate; he describes how the court assessed whether the reports of an informant were reliable in Khan v. Obama. In other words, Tatel seems to be pointing to the kind of problems that a report summarizing Pakistani claims about detainees they were turning over to America might have–and arguing such precedents apply in this case.

Finally, as Tatel explains, one of Kennedy’s concerns it that just a small fraction of the Report was provided in unredacted fashion.

The Report’s heavy redactions–portions of only [redacted] out of [redacted] pages are unredacted–make evaluating its reliability more difficult. The unredacted portions nowhere reveal whether the same person [redacted] or whether someone else performed each of these tasks. And because all the other [redacted] in the Report are redacted, the district court was unable to evaluate the accuracy of [redacted] by inquiring into the accuracy of the Report’s [redacted].


“[F]actual errors” in the Report reinforced the district court’s concerns. [followed by a mostly-redacted paragraph laying out the errors]

Even in what he could see, there were factual errors, but the redactions prevented him from checking the rest. If the report covered a bunch of detainees, I can understand why the government redacted it, but also (given that a number of the men that would have been included have since been determined not to be fighters), redacting the material related to other detainees would hide the fact that the government had since determined the information to be unreliable.

One final note: there are even hints in the Gitmo report that the government came to distrust TD-314/00684-02. Most of the 9 reports that cite it do so minimally–just one or two details supported by the report. The biggest exception is al Harbi’s report, which includes a bunch of references to TD-314/00684-02. The exception is Al Harbi’s report, which was also the earliest of these reports, written in 2006 (most of the rest were written in 2008). Mind you, the government did add “new” information in Latif’s report in 2008–the names of the three Arabs he had seen while purportedly with the Taliban, based off that old report. But aside from that, it seems the government had increasing doubts about relying on this document.

That would be consistent with what DOD itself did in Latif’s case. As Kennedy’s ruling explains, DOD did not make the claims they’re now making–and therefore they did not cite the evidence in The Report–in status reviews in 2004 and 2007.

Specifically, the Department of Defense determined in 2004 that Latif “is not known to have participated in combatant/terrorist training,” JE 79 (Joint Task Force Guantanamo Memorandum recommending transfer of Latif for continued detention in another country, dated [redacted] at 5, and respondents determined in 2007 that Latif should be transferred away from Guantanamo Bay “subject to the process for making appropriate diplomatic arrangements for his departure,” Pet’r’s Mot. to Set Hearing Dates for ISN 156 [], Ex. A (email from Department of Defense employee to counsel for Latif (Feb. 22, 2007)) at 1 [#728], which was not completed.

In any case, it seems like they’ve got just one thing–aside from generalized patterns of travel and the claim that the man who sent Latif for humanitarian care is actually the Al Qaeda recruiter who sent about 7 other Yemenis to train with al Qaeda–that supports their claim Latif was a Taliban fighter. If that one thing is the same bogus claims Pakistanis used to get a bounty for turning Latif over, then this is not just a judicial mistake, but an embarrassment as well.

10 replies
  1. WilliamOckham says:

    Way back when, I started writing some code to process all of those gitmo files, trying to piece together the referenced back docs (like TD-314/00684-02) that appear in multiple files. It was clear that a few key docs (like this one) got cited in many different cases.For various reasons, I never finished that project. Maybe I need to pull that thing off the backburner.

  2. MadDog says:

    With this latest ruling, it reinforces my thought that the lower courts are continuing to chip away at the Supreme Court’s detainee habeas corpus rulings (Rasul v. Bush and Boumediene v. Bush), and that the Supreme Court regrets its rulings and is willingly allowing the lower courts to have their sway.

  3. William Ockham says:

    @emptywheel: Back when McClatchy did their big series on detainees, they published a whole bunch of Gitmo files as pdfs. I downloaded them all and wrote an app to extract all of the supporting doc references and the “facts” cited from each one because I noticed that some docs were cited in multiple files. I suspected that there were a few documents like this one that were foundational to the whole system. I still have the files and the code.

  4. emptywheel says:

    @William Ockham: How hard would it be to check my math here? Like I said, I just googled off the WL set, not the MC set. But I was only going to get stuff that would come up on a text search. The earliest of these–al Harbi–came up on search but then the file didn’t show up in searchees in the doc. I had to eyeball to find it (luckily, that one is almost entirely reliant on that file).

  5. rosalind says:

    OT: ew bait – Farmers markets double, local food sales to hit $7 billion, USDA says

    “Sales made straight from farms to buyers doubled in two decades to $1.2 billion from $650 million in the 1990s, according to the report. Local food sold through intermediary retail channels pulled in $2.7 billion.

    The USDA found that more than 80% of local food producers are small farms making less than $50,000 in gross annual sales. But large farms pulling in $250,000 or more, which represent 5% of the farms with local food offerings, account for in 92% of the associated revenue.”

Comments are closed.