Executive Summary (U) - GlobalSecurity.org...assembled a team of experienced investigators and...

Introduction (U)

(U) On May 25, 2004, Secretary of DefenseDonald H. Rumsfeld directed the Naval InspectorGeneral, Vice Admiral Albert T. Church, III, toconduct a comprehensive review of Department ofDefense (DoD) interrogation operations. Inresponse to this tasking, Vice Admiral Churchassembled a team of experienced investigators andsubject matter experts in interrogation and deten-tion operations. The Secretary specified that theteam was to have access to all documents, records,personnel and any other information deemed rel-evant, and that all DoD personnel must cooperatefully with the investigation. Throughout theinvestigation - which included over 800 interviewswith personnel serving or having served in Iraq,Afghanistan and Guantanamo Bay, Cuba, and sen-ior policy makers in Washington, as well as reviewand analysis of voluminous documentary material- an impressive level of cooperation was evidentthroughout DoD.

(U) Any discussion of military interroga-tion must begin with its purpose, which is to gainactionable intelligence in order to safeguard thesecurity of the United States. Interrogation isoften an adversarial endeavor. Generally,detainees are not eager to provide information,and they resist interrogation to the extent thattheir personal character or training permits.Confronting detainees are interrogators, whosemission is to extract useful information as quickly

as possible. Military interrogators are trained touse creative means of deception and to play upondetainees' emotions and fears even when conduct-ing interrogations of Enemy Prisoners of War(EPWs), who enjoy the full protections of theGeneva Conventions. Thus, people unfamiliarwith military interrogations might view a perfect-ly legitimate interrogation of an EPW, in full com-pliance with the Geneva Conventions, as offensiveby its very nature.

(U) The natural tension that often existsbetween detainees and interrogators has been ele-vated in the post-9/11 world. In the Global War onTerror, the circumstances are different than thosewe have faced in previous conflicts. Human intel-ligence, or HUMINT - of which interrogation is anindispensable component - has taken on increasedimportance as we face an enemy that blends inwith the civilian population and operates in theshadows. And as interrogation has taken onincreased importance, eliciting useful informationhas become more challenging, as terrorists andinsurgents are frequently trained to resist tradi-tional U.S. interrogation methods that aredesigned for EPWs. Such methods - outlined inArmy Field Manual (FM) 34-52, IntelligenceInterrogation, which was last revised in 1992 –have at times proven inadequate in the GlobalWar on Terror; and this has led commanders,working with policy makers, to search for newinterrogation techniques to obtain critical intelli-gence.

1

UNCLASSIFIED • Executive Summary

UNCLASSIFIED

Executive Summary (U)

(U) Interrogation is constrained by legallimits. Interrogators are bound by U.S. laws,including U.S. treaty obligations, and Executive(including DoD) policy - all of which are intended toensure the humane treatment of detainees. Thevast majority of detainees held by U.S. forces dur-ing the Global War on Terror have been treatedhumanely. However, as of September 30, 2004,DoD investigators had substantiated 71 cases ofdetainee abuse, including six deaths. Of note, only20 of the closed, substantiated abuse cases - lessthan a third of the total - could in any way be con-sidered related to interrogation, using broad crite-ria that encompassed any type of questioning(including questioning by non-military-intelligencepersonnel at the point of capture), or any presenceof military-intelligence interrogators. Another 130cases remained open as of September 30, 2004, withinvestigations ongoing.

(U) The events at Abu Ghraib have becomesynonymous with the topic of detainee abuse. Wedid not directly investigate those events, whichhave been comprehensively examined by otherofficials and are the subject of ongoing investiga-tions to determine criminal culpability. Instead,we considered the findings, conclusions and rec-ommendations of previous Abu Ghraib investiga-tions as we examined the larger context ofinterrogation policy development and implemen-tation in the Global War on Terror. In accordancewith our direction from the Secretary of Defense,our investigation focused principally on: (a) thedevelopment of approved interrogation policy

(specifically, lists of authorized interrogation tech-niques), (b) the actual employment of interroga-tion techniques, and (c) what role, if any, theseplayed in the aforementioned detainee abuses. Inaddition, we investigated DoD’s use of civiliancontractors in interrogation operations, DoD sup-port to or participation in the interrogation activ-ities of other government agencies (OGAs), andmedical issues relating to interrogations. Finally,we summarized and analyzed detention-relatedreports and working papers submitted to DoD bythe International Committee of the Red Cross(ICRC). Our primary observations and findingson these issues are set forth below.

(U) Many of the details underlying ourconclusions remain classified, and therefore can-not be presented in this unclassified executivesummary. In addition, we have omitted fromthis summary any discussion of ICRC matters inorder to respect ICRC concerns, and complywith DoD policy, regarding limitation of the dis-semination of ICRC-provided information.Issues of senior official accountability wereaddressed by the Independent Panel to ReviewDoD Detention Operations (hereinafter“Independent Panel”) - chaired by theHonorable James R. Schlesinger - with which weworked closely. Finally, we have based our con-clusions primarily on the information availableto us as of September 30, 2004. Should addi-tional information become available, our conclu-sions would have to be considered in light of thatinformation.

2

UNCLASSIFIED


Interrogation Policy Development (U)

(U) Overview

(U) An early focus of our investigation wasto determine whether DoD had promulgatedinterrogation policies or guidance that directed,sanctioned or encouraged the abuse of detainees.We found that this was not the case. While no uni-versally accepted definitions of “torture” or“abuse” exist, the theme that runs throughout theGeneva Conventions, international law, and U.S.military doctrine is that detainees must be treated“humanely.” Moreover, the President, in hisFebruary 7, 2002 memorandum that determinedthat al Qaeda and the Taliban are not entitled toEPW protections under the Geneva Conventions,reiterated the standard of “humane” treatment.We found, without exception, that the DoD offi-cials and senior military commanders responsiblefor the formulation of interrogation policy evi-denced the intent to treat detainees humanely,which is fundamentally inconsistent with thenotion that such officials or commanders everaccepted that detainee abuse would be permissi-ble. Even in the absence of a precise definition of“humane” treatment, it is clear that none of thepictured abuses at Abu Ghraib bear any resem-blance to approved policies at any level, in anytheater. We note, therefore, that our conclusion isconsistent with the findings of the IndependentPanel, which in its August 2004 report determinedthat “[n]o approved procedures called for or

allowed the kinds of abuse that in fact occurred.There is no evidence of a policy of abuse promul-gated by senior officials or military authorities.”

(U) Nevertheless, with the clarity of hind-sight we consider it a missed opportunity that nospecific guidance on interrogation techniques wasprovided to the commanders responsible forAfghanistan and Iraq, as it was to the U.S.Southern Command (SOUTHCOM) for use atGuantanamo Bay. As the Independent Panelnoted, “[w]e cannot be sure how the number andseverity of abuses would have been curtailed hadthere been early and consistent guidance fromhigher levels.”

(U) Another missed opportunity that weidentified in the policy development process isthat we found no evidence that specific detentionor interrogation lessons learned from previousconflicts (such as those from the Balkans, or eventhose from earlier conflicts such as Vietnam) wereincorporated into planning for operations in sup-port of the Global War on Terror. For example, nolessons learned from previous conflicts were refer-enced in the operation orders (OPORDs) foreither Operation ENDURING FREEDOM (OEF)in Afghanistan or Operation IRAQI FREEDOM(OIF). These OPORDs did cite military doctrineand Geneva Convention protections, but they didnot evidence any specific awareness of the risk ofdetainee abuse - or any awareness that U.S. forceshad confronted this problem before. Though we

3


UNCLASSIFIED

did not find evidence that this failure to highlightthe inherent risk led directly to any detaineeabuse, we recommend that future planning fordetention and interrogation operations in theGlobal War on Terror take full advantage of priorand ongoing experience in these areas.

(U) Set forth below is a brief discussion ofthe significant events in the development of inter-rogation policy for Guantanamo Bay, Afghanistanand Iraq.

(U) Guantanamo Bay, Cuba (GTMO)

(U) Interrogation policy for GTMO hasbeen the subject of extensive debate among boththe uniformed services and senior DoD policymakers. At the beginning of interrogation opera-tions at GTMO in January 2002, interrogatorsrelied upon the techniques in FM 34-52. InOctober 2002, when those techniques had provenineffective against detainees trained to resistinterrogation, Major General Michael E. Dunlavey- the Commander of Joint Task Force (JTF) 170,the intelligence task force at GTMO at the time -requested that the SOUTHCOM Commander,General James T. Hill, approve 19 counter resist-ance techniques that were not specifically listed inFM 34-52. (This request, and descriptions of the19 techniques, were declassified and released tothe public by the Department of Defense on June22, 2004.) The techniques were broken down intoCategories I, II, and III, with the third category

containing the most aggressive techniques. TheSOUTHCOM Commander forwarded the requestto the Chairman of the Joint Chiefs of Staff,General Richard B. Myers, noting that he wasuncertain whether the Category III techniqueswere legal under U.S. law, and requesting addi-tional legal review. On December 2, 2002, on theadvice of the DoD General Counsel, William J.Haynes II, the Secretary of Defense approved theuse of Category I and II techniques, but only oneof the Category III techniques (which authorizedmild, non-injurious physical contact such as grab-bing, poking in the chest with a finger, and lightpushing). The Secretary’s decision thus excludedthe most aggressive Category III techniques: useof scenarios designed to convince the detaineethat death or painful consequences are imminentfor him and/or his family, exposure to cold weath-er or water, and the use of a wet towel and drip-ping water to induce the misperception ofsuffocation. (Notably, our investigation foundthat even the single Category III techniqueapproved was never put into practice.)

(U) Shortly after the December 2, 2002approval of these counter resistance techniques,reservations expressed by the General Counsel ofthe Department of the Navy, Alberto J. Mora, ledthe Secretary of Defense on January 15, 2003 torescind his approval of all Category II techniquesand the one Category III technique (mild, non-inju-rious physical contact), leaving only Category I tech-niques in effect. The same day, the Secretary

4

UNCLASSIFIED


directed that a working group be established toassess interrogation techniques in the Global Waron Terror, and specified that the group should com-prise experts from the Office of General Counsel ofthe Department of Defense, the Office of the UnderSecretary of Defense for Policy, the military servicesand the Joint Staff.

(U) Following a sometimes contentiousdebate, this working group - led by U.S. Air ForceGeneral Counsel Mary Walker, and reporting tothe DoD General Counsel - produced a series ofdraft reports from January through March 2003,including a March 6, 2003 draft report recom-mending approval of 36 interrogation techniques.As many as 39 techniques had been consideredduring the working group’s review, including“water boarding” (pouring water on a detainee’stoweled face to induce the misperception of suffo-cation), which did appear among the 36 tech-niques in the March 6 draft. Four of the 39techniques were considered unacceptable, howev-er - including water boarding - and were ultimate-ly dropped from the review, leaving 35 techniquesthat the working group recommended for consid-eration by the Secretary of Defense. In late March2003, the Secretary of Defense adopted a morecautious approach, choosing to accept 24 of theproposed techniques, most of which were takendirectly from or closely resembled those in FM 34-52. (The 35 techniques considered were reflectedin the working group’s final report, dated April 3,2003.) The Secretary’s guidance was promulgatedto SOUTHCOM for use at GTMO in an April 16,

2003 memorandum (also declassified in June2004) that remains in effect today.

(U) As this discussion demonstrates, theinitial push for interrogation techniques beyondthose found in FM 34-52 came in October 2002from the JTF-170 Commander who, based onexperiences to that point, believed that counterresistance techniques were needed in order toobtain actionable intelligence from detainees whowere trained to oppose U.S. interrogation methods.In addition, the Secretary of Defense moderatedproposed interrogation policies, cutting back on thenumber and types of techniques that were present-ed by some commanders and senior advisors forconsideration. This was true when the Secretaryrejected the three most aggressive Category IIItechniques that JTF-170 requested, and was laterapparent in the promulgation of the April 16, 2003policy, which included only 24 of the 35 techniquesrecommended for consideration by the workinggroup, and included none of the most aggressivetechniques.

(U) Military department lawyers were pro-vided the opportunity for input during the inter-rogation policy debate, even if that input was notalways adopted. This was evident during thereview of JTF-170’s initial request for counterresistance techniques in the lead-up to theDecember 2, 2002 policy, when service lawyer con-cerns were forwarded to the Joint Staff, and laterin the establishment of the working group inJanuary 2003 that led to the April 16, 2003 policy.

5


UNCLASSIFIED

In the first case, in November 2002 the servicesexpressed serious reservations about approvingthe proposed counter resistance techniques with-out further legal and policy review, and thus theywere uncomfortable with the Secretary’s adoptionof a subset of these techniques on December 2,2002. However, in the aftermath of 9/11, the per-ceived urgency of gaining actionable intelligencefrom particularly resistant detainees - includingMohamed al Kahtani, the “20th hijacker” - thatcould be used to thwart possible attacks on theUnited States, argued for swift adoption of aneffective interrogation policy. (In August 2001Kahtani had been refused entry into the U.S. by asuspicious immigration inspector at Florida'sOrlando International Airport, where the lead9/11 hijacker, Mohamed Atta, was waiting forhim.) This perception of urgency was demon-strated, for example, by the SOUTHCOMCommander’s October 2002 memorandum for-warding the counter resistance techniques forconsideration, which stated, “I firmly believe thatwe must quickly provide Joint Task Force 170counter-resistance techniques to maximize thevalue of our intelligence collection mission.”

(U) Afghanistan

(U) Rather than being the subject ofdebate within the Office of the Secretary ofDefense, interrogation techniques for use inAfghanistan were approved and promulgated bythe senior command in the theater. (Initially, thiswas Combined Joint Task Force 180, or CJTF-180,

subsequently renamed CJTF-76. At present,Combined Forces Command-Afghanistan, or CFC-A, commands operations in Afghanistan, withCJTF-76 as a subordinate command.)

(U) From the beginning of OEF in October2001 until December 2002, interrogators inAfghanistan relied upon FM 34-52 for guidance.On January 24, 2003, in response to a Joint Staffinquiry via U.S. Central Command (CENTCOM),the CJTF-180 Acting Staff Judge Advocate for-warded to the CENTCOM Staff Judge Advocate amemorandum that listed and described the inter-rogation techniques then in use in Afghanistan.Many of these techniques were similar to thecounter resistance techniques that the Secretaryhad approved for GTMO on December 2, 2002;however, the CJTF-180 techniques had beendeveloped independently by interrogators inAfghanistan in the context of a broad reading ofFM 34-52, and were described using different ter-minology.

(U) In addition to these locally developedtechniques, however, the January 24, 2003 memo-randum tacitly confirmed that “migration” ofinterrogation techniques had occurred separately.During December 2002 and January 2003, accord-ing to the memorandum, interrogators hademployed some of the techniques approved by theSecretary of Defense for use at GTMO. Use of theTier II and single Tier III technique ceased, how-ever, upon the Secretary’s rescission of their

6

UNCLASSIFIED


approval for GTMO on January 15, 2003.

(U) CJTF-180 did not receive any responseto its January 24, 2003 memorandum from eitherCENTCOM or the Joint Staff, and interpretedthis silence to mean that the techniques then inuse (which, again, no longer included the tieredGTMO techniques) were unobjectionable to high-er headquarters and therefore could be consideredapproved policy.

(U) On February 27, 2003, the CJTF-180Commander, Lieutenant General Dan K. McNeill,revised the January 24, 2003 techniques by modi-fying or eliminating five “interrogator tactics” notfound in FM 34-52 in response to the investigationof the December 2002 deaths of two detainees atthe Bagram Collection Point. While the abusesleading to the Bagram deaths consisted of violentassaults, rather than any authorized techniques,the CJTF-180 Commander modified or eliminatedthese five tactics as a precaution, out of a generalconcern for detainee treatment. This revised pol-icy remained in effect until March 2004, whenCJTF-180 issued new interrogation guidance.

(U) The March 2004 guidance was notdrafted as carefully as it could have or should havebeen. First, it revived some of the practices thatCJTF-180 had modified or eliminated in February2003, without explanation and without even ref-erencing the February 2003 modifications.Second, some of the techniques in the new guid-ance were based upon an unsigned draft memo-

randum from the Secretary of Defense to CENT-COM (prepared by the Joint Staff) that was sub-stantively identical to the Secretary’s April 16,2003 interrogation policy for GTMO. We found noevidence that the Secretary was ever aware of thisdraft memorandum, which was never approved.

(U) The March 2004 interrogation policyremained in effect until June 2004, when theCENTCOM Commander, General John Abizaid,directed that all interrogations in CENTCOM bestandardized under a single policy. The CFC-ACommander, Lieutenant General David W. Barno,then directed that CJTF-76 adopt the existinginterrogation policy used in Iraq, which had beendeveloped in May 2004. This policy relies almostexclusively on interrogation techniques specifical-ly outlined in FM 34-52, and remains in effecttoday.

(U) Iraq

(U) As in Afghanistan, interrogation policyin Iraq was developed and promulgated by thesenior command in the theater, then CombinedJoint Task Force-7, or CJTF-7. At the inception ofOIF on March 19, 2003, interrogators relied uponFM 34-52 for guidance. In August 2003, amid agrowing insurgency in Iraq, Captain CarolynWood, the commander of Alpha Company, 519thMilitary Intelligence Battalion (A/519), stationedat Abu Ghraib, submitted a draft interrogationpolicy directly to the 205th Military IntelligenceBrigade and the CJTF-7 staff. This draft policy

7


UNCLASSIFIED

was based in part on interrogation techniquesbeing used at the time by units in Afghanistan.On August 18, 2003, the Joint Staff’s Director forOperations (J-3) sent a message requesting thatthe SOUTHCOM Commander provide a team ofexperts in detention and interrogation operationsto provide advice on relevant facilities and opera-tions in Iraq. As a result, from August 31 toSeptember 9, 2003, the Joint Task ForceGuantanamo (JTF-GTMO) Commander, MajorGeneral Geoffrey Miller, led a team to assess inter-rogation and detention operations in Iraq. One ofhis principal observations was that CJTF-7 had“no guidance specifically addressing interrogationpolicies and authorities disseminated to units”under its command.

(U) To rectify this apparent problem, theCJTF-7 Commander, Lieutenant General RicardoSanchez, published the first CJTF-7 interrogationpolicy on September 14, 2003. This policy washeavily influenced by the April 2003 JTF-GTMOinterrogation policy, which MG Miller had provid-ed during his visit, and was also influenced by theA/519 draft policy which, as noted above, con-tained some interrogation techniques in use inAfghanistan. However, LTG Sanchez and his staffwere well aware that the Geneva Conventionsapplied to all detainees in Iraq, and thoroughlyreviewed the CJTF-7 policy for compliance withthe Conventions prior to its approval.

(U) After reviewing the September policy

once it was issued, CENTCOM’s Staff JudgeAdvocate considered it overly aggressive. As aresult, CJTF-7 promulgated a revised policy onOctober 12, 2003 that explicitly superseded theprevious policy. This new policy removed severaltechniques that had been approved in theSeptember 2003 policy, rendering the October2003 policy quite similar to the guidance found inFM 34-52. It should be noted that none of thetechniques contained in either the September orOctober 2003 CJTF-7 interrogation policies wouldhave permitted abuses such as those at AbuGhraib.

(U) On May 13, 2004, CJTF-7 issuedanother revised interrogation policy, whichremains in effect today. The list of approved tech-niques remained identical to the October 2003policy; the principal change from the previous pol-icy was to specify that under no circumstanceswould requests for the use of certain techniquesbe approved. While this policy is explicit in itsprohibition of certain techniques, like the earlierpolicies it contains several ambiguities, which -although they would not permit abuse - couldobscure commanders’ oversight of techniquesbeing employed, and therefore warrant reviewand correction. (The details of these ambiguitiesremain classified, but are discussed in the mainbody of this report.) As noted above, in June 2004this policy was adopted for use in Afghanistan.

(U) Subsequent to the completion of this

8

UNCLASSIFIED


report, we were notified that the Commander,Multi-national Forces Iraq (MNF-I), GeneralGeorge W. Casey, Jr., had approved on January 27,2005 a new interrogation policy for Iraq. This pol-icy approves a more limited set of techniques foruse in Iraq, and also provides additional safe-guards and prohibitions, rectifies ambiguities, and– significantly – requires commanders to conducttraining on and verify implementation of the poli-cy and report compliance to the Commander,MNF-I.

Interrogation Techniques ActuallyEmployed by Interrogators (U)

(U) Guantanamo Bay, Cuba

(U) In GTMO, we found that from thebeginning of interrogation operations to the pres-ent, interrogation policies were effectively dissem-inated and interrogators closely adhered to thepolicies, with minor exceptions. Some of theseexceptions arose because interrogation policy didnot always list every conceivable technique thatan interrogator might use, and interrogators oftenemployed techniques that were not specificallyidentified by policy but nevertheless arguably fellwithin the parameters of FM 34-52. This closecompliance with interrogation policy was due to anumber of factors, including strict command over-sight and effective leadership, adequate detentionand interrogation resources, and GTMO’s securelocation far from any combat zone. And although

conditions at GTMO were initially spartan, rely-ing on improvised interrogation booths and pre-existing detention facilities (Camp X-Ray,constructed in the 1990s to house Cuban andHaitian refugees), these conditions continuouslyimproved over time. The most important devel-opment was establishment in November 2002 of acommand organization that placed detention andintelligence operations under the command of asingle entity, JTF-GTMO, superseding the bifur-cated organization which had at times impededintelligence collection due to lack of proper coordi-nation between interrogators and military police.JTF-GTMO, with its well-developed standardoperating procedures and clear lines of authority,enabled effective coordination.

(U) In light of military police participationin many of the abuses at Abu Ghraib, the rela-tionship between military police (MP) and mili-tary intelligence (MI) personnel has come underscrutiny. Under the GTMO model of MP/MI rela-tions, military police work closely with militaryintelligence in helping to set the conditions forsuccessful interrogations, both by observingdetainees and sharing observations with inter-rogators, and by assisting in the implementationof interrogation techniques that are employedlargely outside the interrogation room (such asthe provision of incentives for cooperation). Whenconducted under controlled conditions, with spe-cific guidance and rigorous command oversight, asat GTMO, this is an effective model that greatly

9


UNCLASSIFIED

enhances intelligence collection and does not leadto detainee abuse. In our view, it is a model thatshould be considered for use in other interroga-tion operations in the Global War on Terror.Current MP and MI doctrine, however, is vague onthe proper relationship between MP and MI units,and accordingly requires revision that spells outthe details of the type of coordination betweenthese units that has proven successful at GTMO.

(U) Finally, we determined that during thecourse of interrogation operations at GTMO, theSecretary of Defense approved specific interrogationplans for two “high-value” detainees who had resis-ted interrogation for many months, and who werebelieved to possess actionable intelligence that couldbe used to prevent attacks against the UnitedStates. Both plans employed several of the counterresistance techniques found in the December 2,2002 GTMO policy, and both sucessfully neutralizedthe two detainees’ resistance training and yieldedvaluable intelligence. We note, however, that theseinterrogations were sufficiently aggressive that theyhighlighted the difficult question of precisely defin-ing the boundaries of humane treatment ofdetainees.

(U) Afghanistan and Iraq

(U) Our findings in Afghanistan and Iraqstand in contrast to our findings in GTMO.Dissemination of interrogation policy was generally

poor, and interrogators fell back on their trainingand experience, often relying on a broad interpreta-tion of FM 34-52. In Iraq, we also found generallypoor unit-level compliance with approved policymemoranda even when those units were aware ofthe relevant memoranda. However, in bothAfghanistan and Iraq, there was significant overlapbetween the techniques contained in approved poli-cy memoranda and the techniques that interroga-tors employed based solely on their training andexperience.

(U) While these problems of policy dissem-ination and compliance were certainly cause forconcern, we found that they did not lead to theemployment of illegal or abusive interrogationtechniques. According to our investigation, inter-rogators clearly understood that abusive practicesand techniques - such as physical assault, sexualhumiliation, terrorizing detainees with unmuz-zled dogs, or threats of torture or death - were atall times prohibited, regardless of whether theinterrogators were aware of the latest policy mem-orandum promulgated by higher headquarters.Thus, with limited exceptions (most of which werephysical assaults, as described below in our dis-cussion of detainee abuse), interrogators did notemploy such techniques, nor did they direct MPsto do so. Significantly, nothing in our investiga-tion of interrogation and detention operations inAfghanistan or Iraq suggested that the chaoticand abusive environment that existed at the Abu

10

UNCLASSIFIED


Ghraib prison in the fall of 2003 was repeatedelsewhere.

(U) Nevertheless, as previously stated, weconsider it a missed opportunity that interroga-tion policy was never issued to the CJTF com-manders in Afghanistan or Iraq, as was done forGTMO. Had this occurred, interrogation policycould have benefited from additional expertiseand oversight. In Iraq, by the time the first CJTF-7 interrogation policy was issued in September2003, two different policies had been thoroughlydebated and promulgated for GTMO, and deten-tion and interrogation operations had been con-ducted in Afghanistan for nearly two years. Yet,CJTF-7 was left to struggle with these issues onits own in the midst of fighting an insurgency. Asa result, the September 2003 CJTF-7 interroga-tion policy was developed, as the CJTF-7 StaffJudge Advocate at the time stated, in an “urgent”fashion. Interrogation policy reflecting the les-sons learned to date in the Global War on Terrorshould have been in place in Iraq long beforeSeptember 2003.

(U) Finally, there has been much specula-tion regarding the notion that undue pressure foractionable intelligence contributed to the abuses atAbu Ghraib, and that such pressure also manifesteditself throughout Iraq. It is certainly true that“pressure” was applied in Iraq through the chain ofcommand, but a certain amount of pressure is to beexpected in a combat environment. As LTG

Sanchez has stated, “if I had not been applyingintense pressure on the intelligence community toknow my enemy I would have been derelict in myduties and I shouldn't have been a commandinggeneral.” Our investigation indicated that inter-rogators in Iraq indeed were under intense pressurefor intelligence, but this derived chiefly from a chal-lenging detainee to interrogator (and interpreter)ratio and an inherent desire to help prevent coali-tion casualties. We agree with MG Fay’s observa-tion that pressure for intelligence “should havebeen expected in such a critical situation,” and thatit was not properly managed by unit-level leaders atAbu Ghraib. We found no evidence, however, thatinterrogators in Iraq believed that any pressure forintelligence subverted their obligation to treatdetainees humanely in accordance with the GenevaConventions, or otherwise led them to apply prohib-ited or abusive interrogation techniques. Andalthough Major General Fay’s investigation of theevents at Abu Ghraib noted that requests for infor-mation were at times forwarded directly from vari-ous military commands and DoD agencies to AbuGhraib, rather than through normal channels, wefound no evidence to support the notion that theOffice of the Secretary of Defense, the NationalSecurity Council staff, CENTCOM, or any otherorganization applied explicit pressure for intelli-gence, or gave “back-channel” permission to forcesin the field in Iraq (or in Afghanistan) to use moreaggressive interrogation techniques than thoseauthorized by either command interrogation poli-cies or FM 34-52.

11


UNCLASSIFIED

Detainee Abuse (U)

(U) Overview

(U) We examined the 187 DoD investiga-tions of alleged detainee abuse that had beenclosed as of September 30, 2004. Of these inves-tigations, 71 (or 38%) had resulted in a findingof substantiated detainee abuse, including sixcases involving detainee deaths. Eight of the 71cases occurred at GTMO, all of which were rela-tively minor in their physical nature, althoughtwo of these involved unauthorized, sexuallysuggestive behavior by interrogators, whichraises problematic issues concerning culturaland religious sensitivities. (As described below,we judged that one other substantiated incidentat GTMO was inappropriate but did not consti-tute abuse. This incident was discarded fromour statistical analysis, as reflected in the chartbelow.) Three of the cases, including one deathcase, were from Afghanistan, while the remain-ing 60 cases, including five death cases, occurredin Iraq. Additionally, 130 cases remained open,with investigations ongoing. Finally, our inves-tigation indicated that commanders are makingvigorous efforts to investigate every allegation ofabuse - regardless of whether the allegations aremade by DoD personnel, civilian contractors,detainees, the International Committee of theRed Cross, the local populace, or any othersource.

(U) Included among the open cases wereseveral ongoing investigations related to abuseat Abu Ghraib, including the death of a detaineewho was brought to Abu Ghraib by a specialoperations/OGA team in November 2003.Though not included in our abuse analysis, thiscase was considered in our review of medicalissues. Similarly, the open cases include theDecember 2002 Bagram Collection Point deaths,as those investigations were not completed untilOctober 2004; however, observations on theBagram deaths are provided in our discussionbelow.

(U) We also reviewed a July 14, 2004 let-ter from an FBI official notifying the ArmyProvost Marshal General of several instances of“aggressive interrogation techniques” reported-ly witnessed by FBI personnel at GTMO inOctober 2002. One of these was already the sub-ject of a criminal investigation, which remainsopen. The U.S. Southern Command and the cur-rent Naval Inspector General are now reviewingall of the FBI documents released to theAmerican Civil Liberties Union (ACLU) - which,other than the letter noted above, were notknown to DoD authorities until the ACLU pub-lished them in December 2004 - to determinewhether they bring to light any abuse allega-tions that have not yet been investigated.

(U) For the purposes of our analysis, wecategorized the substantiated abuse cases as

12

UNCLASSIFIED


deaths, serious abuse, or minor abuse. We consid-ered serious abuse to be misconduct resulting orhaving the potential to result in death, or in griev-ous bodily harm (as defined in the Manual forCourts-Martial, 2002 edition.) In addition, weconsidered all sexual assaults, threats to inflictdeath or grievous bodily harm, and maltreatmentlikely to result in death or grievous bodily harm tobe serious abuse. Finally, as noted above, we con-cluded that one of the 71 cases did not constituteabuse for our purposes: this case involved a sol-dier at GTMO who dared a detainee to throw acup of water on him, and after the detainee com-plied, reciprocated by throwing a cup of water onthe detainee. (The soldier was removed from hisassignment as a consequence of inappropriateinteraction with a detainee.) We discarded this

investigation, leaving us 70 substantiateddetainee abuse cases to analyze. The chart belowreflects the breakdown of these 70 abuse cases.

(U) There are approximately 121 abusevictims in these 70 cases of detainee abuse. As ofSeptember 30, 2004, disciplinary action had beentaken against 115 service members for this mis-conduct, including numerous nonjudicial punish-ments, 15 summary courts-martial, 12 specialcourts-martial and nine general courts-martial.

(U) No Connection Between InterrogationPolicy and Abuse

(U) We found no link between approvedinterrogation techniques and detainee abuse. Of

13


UNCLASSIFIED

Closed Substantiated Abuse Cases

27

29

26

1

5

0

10

20

30

40

50

60

70

Afghanistan GTMO Iraq

DeathsSerious AbuseMinor Abuse

Closed Substantiated Abuse Cases (U)

UNCLASSIFIED

the 70 cases of closed, substantiated abuse, only20 of these cases, or less than one-third, could beconsidered “interrogation-related;” the remaining50 were unassociated with any kind of question-ing, interrogation, or the presence of MI person-nel. In determining whether a case wasinterrogation-related, we took an expansiveapproach: for example, if a soldier slapped adetainee for refusing to answer a question at thepoint of capture, we categorized that misconductas interrogation-related abuse - even though it didnot occur at a detention facility, the soldier wasnot an MI interrogator, and there was no indica-tion the soldier was (or should have been) aware ofinterrogation policy approved for use by MI inter-rogators.

(U) At GTMO, where there have been over24,000 interrogation sessions since the beginningof interrogation operations, there are only threecases of closed, substantiated interrogation-relatedabuse, all consisting of minor assaults in which MIinterrogators exceeded the bounds of approvedinterrogation policy. As noted above, these casesincluded those of two female interrogators who, ontheir own initiative, touched and spoke todetainees in a sexually suggestive manner in orderto incur stress based on the detainees’ religiousbeliefs. All three cases resulted in disciplinaryaction against the interrogators.

(U) In Afghanistan, one case of interroga-

tion-related abuse had been substantiated prior to

September 30, 2004. On March 18, 2004, when ele-

ments of a U.S. infantry battalion conducted a cor-

don and search operation in the village of Miam

Do, the U.S. forces were met with resistance and

several Afghans were killed in subsequent fighting.

The unit then detained the entire population of the

village for four days in order to conduct screening

operations. An Army Lieutenant Colonel attached

to the Defense Intelligence Agency accompanied

the battalion during the screening operations, in

which he punched, kicked, grabbed and choked

numerous villagers. As a result, he was disci-

plined and suspended from participating in oper-

ations involving detainees.

(U) In addition, there are now two cases of

closed, substantiated interrogation-related abuse

involving two detainees who died on December 4

and December 10, 2002 at the Bagram Collection

Point in Afghanistan. Those investigations were

not closed until October 2004, after our data analy-

sis had been completed, and thus are not included

in our statistics. We did, however, review the final

Army Criminal Investigative Division (CID)

Reports of Investigation, which included approxi-

mately 200 interviews. We found both investiga-

tions to be thorough in addressing the practices

and leadership problems that led to the deaths and

we note that CID officials have already recom-

mended charges against 15 soldiers (11 MP and

four MI) in relation to the December 4 death, and

27 soldiers (20 MP and seven MI) in relation to the

14

UNCLASSIFIED


December 10 death. (Some of the same personnel

are named in the detention and interrogation of

both detainees.) Significantly, our review of the

investigations showed that while this abuse

occurred during interrogations, it was unrelated to

approved interrogation techniques.

(U) In Iraq, there are 16 cases of closed,substantiated interrogation-related abuse. Five ofthese cases involved MI interrogators. There is nodiscernible pattern in the 16 cases: the incidentsoccurred at different locations and were commit-ted by members of different units. The abusivebehavior varied significantly among these inci-dents, although each involved methods of mal-treatment that were clearly in violation of U.S.military doctrine and U.S. law of war obligations,as well as U.S. interrogation policy. The mostcommon type of detainee abuse was straightfor-ward physical abuse, such as slapping, punchingand kicking. In addition, threats were made innine of the 16 incidents.

(U) As the preceding discussion illustrates,there is no link between any authorized interro-gation techniques and the actual abuses describedin the closed, substantiated interrogation-relatedabuse cases. First, much of the abuse involved thesort of straightforward physical violence thatplainly transgressed the bounds of any interroga-tion policy in any theater, and also violated anydefinition of “humane” detainee treatment.Second, much of the abuse is wholly unconnected

to any interrogation technique or policy, as it wascommitted by personnel who were not MI inter-rogators, and who almost certainly did not know(and had no reason to know) the details of suchpolicy. Nevertheless, these personnel either knewor should have known that their actions wereimproper because they clearly violated militarydoctrine and law of war obligations. And third,even when MI interrogators committed the abuse,their actions were unrelated to any approved tech-niques. Even if interrogators were “confused” bythe issuance of multiple interrogation policieswithin a short span of time, as some have hypoth-esized regarding Abu Ghraib, it is clear that noneof the approved policies - no matter which versionthe interrogators followed - would have permittedthe types of abuse that occurred.

(U) Underlying Reasons for Abuse

(U) If approved interrogation policy did notcause detainee abuse, the question remains: whatdid? While we cannot offer a definitive answer, westudied the DoD investigation reports for all 70cases of closed, substantiated detainee abuse to seeif we could detect any patterns or underlying expla-nations. Our analysis of these 70 cases showed thatthey involved abuses perpetrated by a variety ofactive duty, reserve and national guard personnelfrom three services on different dates and in differ-ent locations throughout Afghanistan and Iraq, aswell as a small number of cases at GTMO. Whilethis diversity argues against a single, overarching

15


UNCLASSIFIED

reason for abuse, we did identify several factors thatmay help explain why the abuse occurred.

(U) First, 23 of the abuse cases, roughly one

third of the total, occurred at the point of capture in

Afghanistan or Iraq - that is, during or shortly after

the capture of a detainee. This is the point at which

passions often run high, as service members find

themselves in dangerous situations, apprehending

individuals who may be responsible for the death or

serious injury of fellow service members. Because of

this potentially volatile situation, this is also the

point at which the need for military discipline is

paramount in order to guard against the possibility

of detainee abuse, and that discipline was lacking in

some instances. Additionally, the nature of the

enemy, and the tactics it has employed in Iraq (and

to a lesser extent, in Afghanistan) may have played

a role in this abuse. Our service members may have

at times permitted the enemy’s treacherous tactics

and disregard for the law of war - exemplified by

improvised explosive devices and suicide bombings -

to erode their own standards of conduct. (Although

we do not offer empirical data to support this con-

clusion, a consideration of past counter-insurgency

campaigns – for example, in the Philippines and

Vietnam – suggests that this factor may have con-

tributed to abuse.) The highly publicized case

involving an Army Lieutenant Colonel in Iraq pro-

vides an example. On August 20, 2003, during the

questioning of an Iraqi detainee by field artillery sol-

diers, the Lieutenant Colonel fired his weapon near

the detainee's head in an effort to elicit information

regarding a plot to assassinate U.S. service mem-

bers. For his actions, the Lieutenant Colonel was

disciplined and relieved of command.

(U) Second, there was a failure to react to

early warning signs of abuse. Though we cannot

provide details in this unclassified executive sum-

mary, it is clear that such warning signs were pres-

ent - particularly at Abu Ghraib - in the form of

communiqués to local commanders, that should

have prompted those commanders to put in place

more specific procedures and direct guidance to pre-

vent further abuse. Instead, these warning signs

were not given sufficient attention at the unit level,

nor were they relayed to the responsible CJTF com-

manders in a timely manner.

(U) Finally, a breakdown of good order and

discipline in some units could account for other inci-

dents of abuse. This breakdown implies a failure of

unit-level leadership to recognize the inherent

potential for abuse due to individual misconduct, to

detect and mitigate the enormous stress on our

troops involved in detention and interrogation oper-

ations, and a corresponding failure to provide the

requisite oversight. As documented in previous

reports (including MG Fay’s and MG Taguba’s

investigations), stronger leadership and greater

oversight would have lessened the likelihood of

abuse.

16

UNCLASSIFIED


Use of Contract Personnel inInterrogation Operations (U)

(U) It is clear that contract interrogatorsand support personnel are “bridging gaps” in theDoD force structure in GTMO, Afghanistan andIraq. As a senior intelligence officer at CENT-COM stated: “[s]imply put, interrogation opera-tions in Afghanistan, Iraq and Guantanamocannot be reasonably accomplished without con-tractor support.” As a result of these shortfalls incritical interrogation-related skills, numerouscontracts have been awarded by the services andvarious DoD agencies. Unfortunately, however,this has been done without central coordination,and in some cases, in an ad hoc fashion (as demon-strated, for example, by the highly publicized useof a “Blanket Purchase Agreement” administeredby the Department of the Interior to obtain inter-rogation services in Iraq from CACI, Inc.).Nevertheless, we found - with limited exceptions -that contractor compliance with DoD policies, gov-ernment command and control of contractors, andthe level of contractor experience were satisfacto-ry, thanks in large part to the diligence of con-tracting officers and local commanders.

(U) Overall, we found that contractors

made a significant contribution to U.S. intelligence

efforts. Contract interrogators were typically for-

mer MI or law enforcement personnel, and on

average were older and more experienced than mil-

itary interrogators; many anecdotal reports indi-

cated that this gave contract interrogators addi-

tional credibility in the eyes of detainees, thus pro-

moting successful interrogations. In addition,

contract personnel often served longer tours than

DoD personnel, creating continuity and enhancing

corporate knowledge at their commands.

(U) Finally, notwithstanding the highlypublicized involvement of some contractors inabuse at Abu Ghraib, we found very few instancesof abuse involving contractors. In addition, a com-prehensive body of federal law permits the prose-cution of U.S. nationals - whether contractor,government civilian, or military - who may beresponsible for the inhumane treatment ofdetainees during U.S. military operations over-seas. Thus, contractors are no less legallyaccountable for their actions than their militarycounterparts.

DoD Support to Other GovernmentAgencies (U)

(U) For the purposes of our discussion,other government agencies, or OGAs, are federalagencies other than DoD that have specific inter-rogation and/or detention-related missions in theGlobal War on Terror. These agencies include theCentral Intelligence Agency (CIA), the FederalBureau of Investigation (FBI), the DrugEnforcement Administration (DEA), U.S.Customs and Border Protection, and the SecretService. In conducting our investigation, we con-

17


UNCLASSIFIED

sidered DoD support to all of these agencies, butwe focused primarily on DoD support to the CIA.(The CIA cooperated with our investigation, butprovided information only on activities in Iraq.) Itis important to highlight that it was beyond thescope of our tasking to investigate the existence,location or policies governing detention facilitiesthat may be exclusively operated by OGAs, ratherthan by DoD.

(U) DoD personnel frequently workedtogether with OGAs to support their commonintelligence collection mission in the Global Waron Terror, a cooperation encouraged by DoD lead-ership early in Operation ENDURINGFREEDOM. In support of OGA detention andinterrogation operations, DoD provided assistancethat included detainee transfers, logistical func-tions, sharing of intelligence gleaned from DoDinterrogations, and oversight and support of OGAinterrogations at DoD facilities. However, wewere unable to locate formal interagency proce-dures that codified the support roles and process-es.

(U) In OEF and OIF, senior military com-manders were issued guidance that required notifi-cation to the Secretary of Defense prior to thetransfer of detainees to or from other federal agen-cies. This administrative transfer guidance was fol-lowed, with the notable exception of occasions whenDoD temporarily held detainees for the CIA - includ-ing the detainee known as “Triple-X” - without

properly registering them and providing notifica-tion to the International Committee of the RedCross. This practice of holding “ghost detainees”for the CIA was guided by oral, ad hoc agreementsand was the result, in part, of the lack of any specif-ic, coordinated interagency guidance. Our reviewindicated, however, that this procedure was limitedin scope. To the best of our knowledge, there wereapproximately 30 “ghost detainees,” as compared toa total of over 50,000 detainees in the course of theGlobal War on Terror. The practice of DoD holding“ghost detainees” has now ceased.

(U) Aside from the general requirement totreat detainees humanely, we found no specificDoD-wide direction governing the conduct of OGAinterrogations in DoD interrogation facilities. Inresponse to questions and interviews for ourreport, however, senior officials expressed clearexpectations that DoD-authorized interrogationpolicies would be followed during any interroga-tion conducted in a DoD facility. For example, theJoint Staff J-2 stated that “[o]ur understanding isthat any representative of any other governmen-tal agency, including CIA, if conducting interroga-tions, debriefings, or interviews at a DoD facilitymust abide by all DoD guidelines.” On many occa-sions, DoD and OGA personnel did conduct jointinterrogations at DoD facilities using DoD-authorized interrogation techniques. However,our interviews with DoD personnel assigned tovarious detention facilities throughoutAfghanistan and Iraq demonstrated that they did

18

UNCLASSIFIED


not have a uniform understanding of what rulesgoverned the involvement of OGAs in the interro-gation of DoD detainees. Such uncertainty couldcreate confusion regarding the permissibility andlimits of various interrogation techniques. Wetherefore recommend the establishment and widepromulgation of interagency policies governingthe involvement of OGAs in the interrogation ofDoD detainees.

Medical Issues Related toInterrogation (U)

(U) In reviewing the performance of med-ical personnel in detention and interrogation-related operations during the Global War onTerror, we were able to draw preliminary insightsin four areas: detainee screening and medicaltreatment; medical involvement in interrogation;interrogator access to medical information; andthe role of medical personnel in preventing andreporting detainee abuse. We note that the Officeof the Secretary of Defense is currently developingspecific policies to address all of the issues raisedbelow.

(U) First, the medical personnel that weinterviewed understood their responsibility toprovide humane medical care to detainees, inaccordance with U.S. military medical doctrineand the Geneva Conventions. The essence ofthese requirements is captured succinctly in aDoD policy issued by the Assistant Secretary of

Defense for Health Affairs on April 10, 2002,“DoD Policy on Medical Care for Enemy PersonsUnder U.S. Control Detained in Conjunction withOperation Enduring Freedom.” The policy states,“[i]n any case in which there is uncertainty aboutthe need, scope, or duration of medical care for adetainee under U.S. control, medical personnelshall be guided by their professional judgmentsand standards similar to those that would be usedto evaluate medical issues for U.S. personnel, con-sistent with security, public health management,and other mission requirements” (emphasisadded). Few U.S. personnel, however, hadreceived specific training relevant to detaineescreening and medical treatment. As a result, inAfghanistan and Iraq we found inconsistent field-level implementation of specific requirements,such as monthly detainee inspections and weightrecordings. Thus there is a need for a focusedtraining program in this area so that our medicalpersonnel are aware of and comply with detaineescreening and medical treatment requirements.

(U) Second, it is a growing trend in theGlobal War on Terror for behavioral science person-nel to work with and support interrogators. Thesepersonnel observe interrogations, assess detaineebehavior and motivations, review interrogationtechniques, and offer advice to interrogators. Thissupport can be effective in helping interrogators col-lect intelligence from detainees; however, it must bedone within proper limits. We found that behavioralscience personnel were not involved in detainee

19


UNCLASSIFIED

medical care (thus avoiding any inherent conflictbetween caring for detainees and crafting interroga-tion strategies), nor were they permitted access todetainee medical records for purposes of developinginterrogation strategies. However, since neither theGeneva Conventions nor U.S. military medical doc-trine specifically address the issue of behavioral sci-ence personnel assisting interrogators in developinginterrogation strategies, this practice has evolved inan ad hoc manner. In our view, DoD policy-levelreview is needed to ensure that this practice is per-formed with proper safeguards, as well as to clarifythe status of medical personnel (such as behavioralscientists supporting interrogators) who do not par-ticipate in patient care.

(U) Another area that deserves DoD poli-cy-level review (and that is unaddressed by theGeneva Conventions or current DoD policy) isinterrogator access to detainee medical informa-tion. Interrogators often have legitimate reasonsfor inquiring into detainees’ medical status. Forexample, interrogators need to be able to verifywhether detainees are being truthful when theyclaim that interrogations should be restricted onmedical grounds. Granting interrogators unfet-tered access to detainee medical records, however,raises the problem that detainee medical informa-tion could be inappropriately exploited duringinterrogations. Such access might also discouragedetainees from being truthful with medical per-sonnel, or from seeking help with medical issues,if detainees believe that their medical histories

will be used against them during interrogation.Although U.S. law provides no absolute confiden-tiality of medical information for any person,including detainees, DoD policy-level review isnecessary in order to balance properly these com-peting concerns. This is especially true given thesubstantial variation that we found in field-levelpractices for maintaining and securing detaineemedical records. While access to medical informa-tion was carefully controlled at GTMO, we foundin Afghanistan and Iraq that interrogators some-times had easy access to such information.Nevertheless, we found no instances wheredetainee medical information had been inappro-priately used during interrogations, and in mostsituations interrogators had little interest indetainee medical information even when they hadunfettered access to it.

(U) Finally, it was not possible for us toassess comprehensively whether medical personnelserving in the Global War on Terror have adequate-ly discharged their obligation to report (and wherepossible, prevent) detainee abuse. However, ourinterviews with medical personnel indicated thatthey had only infrequently suspected or witnessedabuse, and had in those instances reported itthrough the chain of command. Separately, we per-formed a systematic review of investigative notesand autopsy results in order to assess the roles ofmedical personnel, especially in any case wheredetainee abuse was suspected. We reviewed 68detainee deaths: 63 in Iraq and five in Afghanistan;

20

UNCLASSIFIED


there were no deaths at GTMO. (These deaths werenot all abuse-related, and therefore do not correlatedirectly to the death cases described in our analysisof abuse.) Of these deaths, we identified three inwhich it appeared that medical personnel may haveattempted to misrepresent the circumstances ofdeath, possibly in an effort to disguise detaineeabuse. Two of these were the previously describeddeaths in Bagram, Afghanistan in December 2002,and one was the aforementioned death at AbuGhraib in November 2003. The Army SurgeonGeneral is currently reviewing the specific medicalhandling of these three cases.

Conclusion (U)

(U) Human intelligence in general, andinterrogation in particular, are indispensable com-ponents of the Global War on Terror. The need forintelligence in the post-9/11 world, and ourenemy’s ability to resist interrogation, havecaused our senior policy makers and military com-manders to reevaluate traditional U.S. interroga-tion methods and search for new and moreeffective interrogation techniques. According toour investigation, this search has always been

conducted within the confines of our armedforces’ obligation to treat detainees humanely. Inaddition, our analysis of 70 substantiated detaineeabuse cases found that no approved interrogationtechniques caused these criminal abuses; however,two specific interrogation plans approved foruse at GTMO did highlight the difficulty of pre-cisely defining the boundaries of humane treat-ment.

(U) It bears emphasis that the vast major-ity of detainees held by the U.S. in the Global Waron Terror have been treated humanely, and thatthe overwhelming majority of U.S. personnel haveserved honorably. For those few who have not,there is no single, overarching explanation. Whileauthorized interrogation techniques have notbeen a causal factor in detainee abuse, we havenevertheless identified a number of missed oppor-tunities in the policy development process. Wecannot say that there would necessarily have beenless detainee abuse had these opportunities beenacted upon. These are opportunities, however,that should be considered in the development offuture interrogation policies.

21


UNCLASSIFIED

Executive Summary (U) - GlobalSecurity.org...assembled a team of experienced investigators and...

Documents

Transcript of Executive Summary (U) - GlobalSecurity.org...assembled a team of experienced investigators and...