When the Senate Intelligence Committee initially released its Study on the CIA’s Enhanced Interrogation Program in December 2014, the CIA quietly released a Note to the Reader along with its Fact Sheet, statement from Director John Brennan, and June 2013 Response to a draft of the SSCI Study. The Note to the Reader updates the CIA’s June 2013 response to a draft of the SSCI Study.
According to media reports, the Senate Intelligence Committee only became aware of the document last week. The Note itself states that the revisions presented in it came after discussions with the SSCI staff following the CIA’s 2013 response, which was set to be released in full—without changes—in response to a pending Freedom of Information Act request. The Note to the Reader is neither listed in the December 2014 archive of press releases and statements, nor is it available on the CIA Fact Sheet Regarding the SSCI Study.
The Note details 13 corrections to the CIA’s response to the SSCI. While some reflect small changes in language, others modify the characterization of significant factual claims. For instance, the Note corrects the CIA’s characterization that “the Secretary of State, Deputy Secretary of State and Ambassadors in detention site host countries were aware of the sites at the time they were operational,” stating that while Ambassadors were made aware of detention sites, the Agency cannot “verify” that the Secretary of State and Deputy Secretary of State were independently made aware of every site.
In another instance, the Agency corrects the number of detainees the that were determined to have passed through CIA custody as of January 2009. The CIA originally said that the CTC determined there were “as many as 112.” Instead, the CTC’s count at time was “at least 112.”
Perhaps most significantly, the Note corrects the characterization that information from detainee Khalid Sheikh Mohammed was crucial to finding Riduan Isamuddin, known as Hambali, because the information was otherwise unavailable. The Note suggests that the information used to zero in on Hambali had already been obtained by the FBI in its debriefing of detainee Majid Khan.
The sequencing of the information obtained from Khan and KSM is important because it potentially undermines the Agency’s claim that its enhanced interrogation techniques provided otherwise unavailable information. However, the CIA suggests that the main thrust of its revisions are actually to confirm the CIA’s conclusions regarding overstatements made in the SSCI Study.
According to the Note, the CIA acknowledges “that in various representations, including President Bush's 2006 speech, CIA introduced a sequencing error regarding Majid Khan's arrest/debriefings, and KSM's arrest/debriefings.” The Agency concedes that it “should have written that ‘KSM provided information that led us to understand the significance of a Jemaah lslamiya operative named Zubair.’” Instead, the CIA “incorrectly stated that KSM's information preceded Majid Khan's information.” Even so, the Agency argues that its “building block” approach was critical to recognizing the importance of different strands of intelligence, and “the impact of the information acquired from KSM in the Hambali case remains accurate. It was the combination of information from both terrorists [KSM and Khan] that caused us to focus on Zubair as an inroad to Hambali.” The Note concludes that the analysts from the CIA “stand by our overall conclusion regarding the value of KSM’s information.”
At the time of its release, Lawfare compiled a five-part analysis of the findings, conclusions, and areas of dispute between the SSCI Study, the Minority responses, and the CIA’s comments. Lawfare has updated its early coverage of the SSCI Study to include the CIA’s December 2014 Note to the Reader.
Below, the corrections from the December 2014 Note to the Reader are paired with the corresponding section of the CIA’s June 2013 Response that it updates.
Tab A, p. 10, tic 3; and Tab B, p. 44, para 4: We should not have included in our count of 29 Inspector General investigations of misconduct those cases involving detainees in Iraq or actions that occurred during renditions or transfers of detainees because those cases fall outside the scope of the Committee's Study. The two specific examples described in Tab B, p. 44, para 4 were among this subset of cases that are outside the Study's scope. Our records indicate the number of investigations that are within the scope of the Study is 13.
The OIG conducted two separate major reviews and at least 29 separate investigations of allegations of misconduct. Some of these reviews were self-initiated by Agency components responsible for managing the program. CIA made numerous referrals to the GIG relating to the conduct of Agency officers and their treatment of detainees, during the life of the program as well as after. (Tab A, p.10 tic 3)
Accountability Outcomes. CIA's RDI activities engendered a significant number of accountability-related actions. The IG, often in response to CIA referrals, conducted at least 29 investigations of RDI-related conduct, plus two wide-ranging reviews of the program. Many cases were investigated by the IG and found to be without merit. Of the cases which were found to be supported by the facts, one involved the death of an Afghan national who was beaten by a contractor. The individual involved was prosecuted by the Department of Justice and convicted on a felony charge. Another case involved a contractor who slapped, kicked, and struck detainees when they were in military custody. Shortly after the IG concluded its investigation of that case, the contractor was terminated from the CIA, had his security clearances revoked, and was placed on a contractor watchlist. (Tab B, p. 44, para 4)
Tab B, p. 11, para 2, tic 2: Ambassadors were made aware of detention sites when they were operational but we cannot verify from our records that the Secretary and Deputy Secretary of State were made aware of every site from the onset of operations at each site.
Finally, with regard to the Study’s claims that the State Department was "cut out" of information relating to the program, the record shows that the Secretary of State, Deputy Secretary of State and Ambassadors in detention site host countries were aware of the sites at the time they were operational. In addition, Station Chiefs in the respective countries informed their Ambassadors of developing media, legal, or policy issues as they emerged, and provided a secure communication channel for discussion of these matters with Washington. (Tab B, p. 11, para 2, tic 2)
Tab B, p. 19, para 1: Our response correctly noted that the effort by the Counterterrorism Center (CTC) to definitively establish the number of detainees held over the course of the program remained incomplete by the time a CTC officer briefed Director Hayden on that effort. However, we incorrectly characterized the status of CTC's count at that time. CTC's count was at least 112 (not "as high as 112"), even without the inclusion of detainees who were no longer in CIA's custody prior to consolidation of the program in December 2002. Including those earlier detainees would have added to CTC's count.
Third, we disagree with the Study’s implication that statements made by Director Hayden to the effect that CIA held 98 detainees reflected an attempt to misrepresent the scope of the program to the incoming administration. Director Hayden did meet with CTC and other officers in January 2009, to discuss his upcoming briefing to incoming officials. At that meeting, a CTC officer briefed the research he had performed on the number of total detainees through the life of the Program. Although this research, which indicates the total number of detainees could have been as high as 112, is heavily cited by the SSCI Study, SSCI neglects to point out that the findings were not final. As the briefing stated, "these numbers will continue to be refined as methodical reviews of operational records are completed and disparately compartmented information is researched and consolidated." (Tab B, p.19, para 1)
Tab B, p. 37, para 1, tic 2, sentence 2: Interrogation training began in November 2002, not November 2003.
In his statement for the record, DCIA Hayden noted as an example of a safeguard CIA had built into the program that all those involved in the questioning of detainees are carefully selected and trained. We concede that prior to promulgation of DCIA guidance on interrogation in January 2003 and the establishment of interrogator training courses in November of the same year, not every CIA employee who debriefed detainees had been thoroughly screened or had received formal training. After that time, however-the period with which DCIA Hayden, who came to the Agency in 2005, was most familiar-the statement is accurate. (Tab B, p. 37, para 1, tic 2)
Tab B, pp. 57-58 (also referenced on page 47, first partial para): In addressing the Study's allegation of inadequate accountability for 16 alleged cases in which enhanced techniques were used without prior Headquarters approval, the explanation we provided for the miscount was inaccurate or incomplete in some respects. Based on further research, we offer the following refined and amended explanations for the 16 cases:
- In four cases, no techniques categorized at the time as "enhanced" were used.
- In five cases, enhanced techniques were used without Headquarters approval before the January 2003 guidance requiring such approval. All of these techniques had been assessed by the Department of Justice's Office of Legal Counsel to be lawful.
- In four cases, the need for accountability was mitigated because an interrogation plan had been approved and the specific techniques used beyond those listed in the plan did not exceed its overall scope and were subsequently approved by Headquarters.
- In three cases, unapproved techniques were used and the officers involved were referred to the Inspector General.
This breakdown confirms our overall contention that the draft Study overstated both the number of detainees on whom enhanced techniques were used without prior authorization and the extent to which accountability fell short when such authorization was not obtained.
We agree that there were instances in which CIA used inappropriate and unapproved interrogation techniques, particularly at the program’s outset. Overall, however, we believe that the Study overstates the number of instances of unauthorized use of enhanced techniques as well as the number of non-certified individuals whom it alleges wrongfully participated in interrogations. The Study also overlooks the fact that, subsequent to CIA’s efforts to organize and consolidate its detention and interrogation efforts into one Headquarters-managed program, the Agency worked to ensure that allegations of wrongdoing were reported to management, the Office of Inspector General, and/or the Department of Justice (DOJ), as appropriate.
Moreover, while it would have been prudent to seek guidance from OLC on the complete range of techniques prior to their use, we disagree with any implication that, absent prior OLC review, the use of the “unapproved" techniques was unlawful or otherwise violated policy.
The Study’s assertion that 16 detainees were subjected to enhanced techniques without authorization from CIA Headquarters seems founded on a misunderstanding of the facts. The Study arrives at this number largely by conflating standard interrogation techniques that did not require prior approval with enhanced interrogation techniques that did. Some of this confusion is understandable, as over time, the term “standard" techniques was eliminated and some techniques which were initially classified as “standard" eventually were reclassified as “enhanced."
The Study correctly identifies seven instances in which detainees were subjected to individual techniques which were not approved in advance and included in their interrogation plans. In several of these, however, Headquarters had approved interrogation plans for the detainees utilizing other enhanced techniques. For instance, our review of contemporaneous cable traffic indicates that at (REDACTED) Libyans Abu Hazim and 'abd ai-Karim appear to have been subjected to walling without prior approval. Muhammad Umar 'Abd ai-Rahman, also known as “Asadallah," and 'abd ai-Karim appear to have been subjected to cramped confinement without prior Headquarters approval. (REDACTED) detainee Ramzi bin al Shibh appears to have been subjected to the use of the facial hold technique without prior approval. In these cases, other previously approved enhanced techniques were also used.
In the cases involving Abu Hazim and 'abd al-Karim, Headquarters approved the techniques the following month as components of revised interrogation plans. In the case of Ramzi bin al-Shibh, a cable exchange 18 days after he was subjected to the facial hold indicated Headquarters support for the use of the technique so long as necessary medical personnel were on scene.
However, nine of the Study’s examples describe the application, not of enhanced techniques, but of techniques that were classified at the time as standard. The DCI Guidelines for the Conduct of Interrogation, issued in January 2003, explicitly required prior written approval in advance for use of enhanced techniques, but the guidelines did not require such approvals for the use of standard techniques. While sleep deprivation, nudity, bathing, water dousing, and dietary manipulation were later reclassified as enhanced techniques, they were defined as standard techniques not requiring prior approval at the time relevant to the examples cited in the Study. As a consequence, it is misleading to assert that either officers or CIA's management of the RDI program erred by failing to obtain prior written approvals.
We also believe it is important to note that half of the 16 examples cited in the Study concern detainees who were held at (REDACTED) prior to 3 December 2002, before (REDACTED) formal transition to RDG supervision and subsequent imposition, in January 2003, of guidance on standardized program techniques and approval processes for detention and interrogation operations in (REDACTED). The 2004 OIG Special Review catalogued the use of unapproved and inappropriate techniques at (REDACTED) from September through December 2002, and we have acknowledged serious shortcomings at (REDACTED) in several of our responses to Study conclusions. However, after the standard was approved and communicated in January 2003, interrogation operations at (REDACTED) were generally in line with the guidance with some isolated exceptions identified in the Study and described elsewhere in our response.
The Study also asserts that CIA officers employed water dousing even though CIA Headquarters offered no guidance on the technique until January 2004. That is incorrect. We identified several Headquarters cables dated as early as 2 March 2003 which contained clear instruction on conditions required in order to apply water dousing in a safe and sanctioned manner. Subsequent Headquarters-originated cables were also located dating to June 2003, which classified the application of the technique as a "standard" technique. In September 2003, draft OMS guidelines also discussed water dousing as a standard technique and provided guidance to OMS personnel on its safe application.
The Study further asserts that in "at least eight" interrogations, officers participated without approval of CIA Headquarters. We are unable to locate and identify within the Study all eight instances to which the underlying text of Conclusion 20 refers. We presume the allegation is intended to reference interrogations involving non-certified officers. In reaching this conclusion, the Study appears to rely upon information taken out of context and, in other cases, simply fails to provide supporting evidence. (Tab B, pp. 57-58)
Tab C, p. 5, last para: We acknowledge that there was at least one instance after 2007 of a CIA officer validating an earlier document that referenced the "dirty bomb" aspect of Padilla's plotting.
We assess to this day that Padilla was a legitimate threat who had been directed to use his training in Afghanistan, funding from al-Qa’ida, and US passport to put together a plan to attack tall residential buildings. It took us until 2007 to consistently stop referring to his association with the "Dirty Bomb” plot-a plan we concluded early on was never operationally viable. (Tab C, p. 5, last para)
Tab C, p. 9, first full tic, last sentence: "Students' interest" should have been written "student's interest" because we know of only one student in the group who had an interest in aviation.
Hambali, after having undergone enhanced techniques in CIA detention, admitted he had hand-picked these students in response to KSM's request and that some were being groomed as pilots for unspecified al-Qa’ida operations. Hambali did subsequently recant this statement, claiming he made it to satisfy his interrogators and relieve the pressure of enhanced techniques. We continue to assess his original revelation was correct, however, based on KSM's claim that he tasked Hambali to identify and train pilots, Hambali's verification of this claim in multiple instances, and the students' interest in aircraft and aviation.
Tab C, p. 15, bolded summary, last sentence: We should not have said the fragmentary information was "unavailable to CIA." FBI information on a second shoe bomber does appear in a January 2002 briefing item prepared for CIA's daily counterterrorism update. We do not know how the item's author learned this information and it does not appear to have been entered into record traffic that would have been readily available to analysts. Instead, the record shows clearly that information from Khalid Shaykh Muhammad (KSM) was central to CIA's effort to identify the second shoe bomber.
CIA accurately represented that Khalid Shaykh Muhammad's (KSM) information was central to our efforts to identify and enable British liaison to arrest Sajid Badat, an al-Qa'ida operative who originally planned to conduct a shoe bomb attack aboard an airplane. KSM was the first to tell us there was a second shoe bomber and that he remained at large, and he provided sufficient details to allow CIA and British authorities to identify Badat. Fragmentary information implied a second shoe bomber existed before KSM's detention, but this information was either inconclusive or not available to CIA. (Tab C, p. 15, balded summary, last sentence)
Tab C, p. 17, third tic: Rather than "the individual managing the plot," we should have written "the individual who was in a position to advance the plot." This terrorist had raised Canary Wharf as a potential target and was tasked by KSM to conduct surveillance of Heathrow Airport's security, but the plot was shelved after KSM's arrest.
CIA obtained updated information from KSM about the plot to attack Heathrow Airport and Canary Wharf after he had been subjected to enhanced techniques, including the information on the individual managing the plot, Abu Talha ai-Pakistani. (Tab C, p. 17, third tic)
Tab C, p. 19, bolded summary, first sentence: Instead of "KSM provided information on an al-Qa'ida operative named Zubair...," we should have written that "KSM provided information that led us to understand the significance of a Jemaah lslamiya operative named Zubair." We acknowledge that in various representations, including President Bush's 2006 speech, CIA introduced a sequencing error regarding Majid Khan's arrest/debriefings, and KSM's arrest/debriefings. We repeated that error here and on page 26 of Tab C (see next erratum). However, despite that error, our description of the impact of the information acquired from KSM in the Hambali case remains accurate. It was the combination of information from both terrorists that caused us to focus on Zubair as an inroad to Hambali.
CIA accurately cited Khalid Shaykh Muhammad's (KSM) reporting as a crucial link in a chain of events that led to the capture of Hambali. KSM provided information on an al-Qa'ida operative named Zubair, we shared that lead with Thai authorities, they detained Zubair, and he gave actionable information that helped us identify Hambali's location. Although we had some other information linking Zubair to al-Qa'ida's Southeast Asian network, the record shows clearly that it was KSM's information that caused us to focus on him as an inroad to Hambali, so we continue to assess this is a good example of the importance of intelligence derived from detainee reporting in helping to capture other terrorists. (Tab C, p. 19, bolded summary, first sentence)
Tab C, p. 26, italicized tic (also referenced in Tab B, p. 22, first tic, last sentence; and Tab C, p. 2, para 5, first tic, last sentence): In our review of this case, we correctly acknowledged that CIA allowed a mistaken claim that KSM played a role in Majid Khan's capture to appear in the Inspector General's 2004 Special Review, and we correctly wrote that this claim was a one-time error. However, our effort to provide an example of a more accurate "typical representation" of the relationship between KSM's information and Khan ran afoul of the sequencing error noted in the previous erratum. Although information from KSM was used to elicit further details from Khan, by then Khan already had provided the information that, together with what we learned from KSM, enabled us to advance our search for Hambali.
The CIA repeatedly represented that the CIA interrogation program, and/or the CIA’s enhanced interrogation techniques, resulted in critical, otherwise unavailable intelligence, related to...the capture of Majid Khan.
CIA mistakenly provided incorrect information to the Inspector General (lG) that led to a one time misrepresentation of this case in the IG's 2004 Special Review. This mistake was not, as it is characterized in the "Findings and Conclusions" section of the Study, a "repeatedly represented" or "frequently cited" example of the effectiveness of CIA's interrogation program. CIA accurately described the importance of Khalid Shaykh Muhammad's (KSM) information in the Majid Khan case in a number of finished analytic reports and briefings before and after the Special Review.
Broadly disseminated DI finished intelligence, as well as briefings and materials provided to the SSCl, the White House, the Department of Justice, and the American public-both before and after the Special Review-included accurate representations regarding Majid Khan's importance.
The standard language we used to describe Majid Khan did not imply KSM's information played a role in his capture and instead focused on the importance of his information as a building block that led to other operational successes. For example, a typical representation stated:
“KSM provided information about an al-Qa'ida operative, Majid Khan, who he was aware had recently been captured. KSM-possibly believing the detained operatives was “talking” admitted to having tasked Majid with delivering a large sum of money to individuals working for another senior al-Qa'ida associate. In an example of how information from one detainee can be used in debriefing another detainee in a “building block” process, Khan-confronted with KSM’s information about the money-acknowledged that he delivered the money to an operative named Zubair and provided Zubair’s physical description and contact number.” (Full section provided for clarity)
Tab C, p. 32, last sentence: We incorrectly stated that KSM's information preceded Majid Khan's information. We stand by our overall conclusion regarding the value of KSM's information.
The Study cites two other pieces of information on Paracha that it claims are representative of reporting available independent of the CIA detention and interrogation program. Neither report was noteworthy without KSM's information.
- One is an indirect connection to Paracha's business in Pakistan that Committee staff found in an undisseminated FBI case file. It was not available to CIA at the time and would not have linked Paracha to an al-Qa'ida operation independent of KSM's information in any case. The other report is (REDACTED) of Majid Khan before he was rendered (REDACTED) to US custody, but the report included few details and was disseminated just after KSM provided the information that allowed us to identify Paracha. (Tab C, p. 32, last sentence)
Tab C, p. 34, last tic: We should have written "our" instead of "the early" because this error was not confined to some of CIA's early representations.
CIA records indicate that KSM did not know al-Tayar’s true name and that it was Jose Padilla-in military custody and questioned by the FBI-who provided al- Tayar’s true name as “Adnan el-Shukrijumah.”
While KSM did not know al-Tayyar's true name, his biographic description was sufficient for FBI to identify Adnan el-Shukrijumah as a likely candidate. In addition, the FBI knew to ask Padilla about al-Tayyar's true name because KSM told CIA debriefers that he would know it.
In reviewing this case, we did identify occasions when CIA's language either was not as precise as it should have been or we made factual errors.
- Sometimes we said KSM called al-Tayyar the “next Muhammad Atta." This was an imprecise paraphrase of KSM, who actually described al-Tayyar as having similar education and Western experience as Muhammad Atta and considered him as the “next emir" for an attack in the United States. KSM did not call al-Tayyar “the next Muhammad Atta."
- In some of the early representations, we incorrectly stated al-Tayyar fled the United States in response to the FBI investigation, although he had in fact already departed the United States by this time. (Full section provided for context)