Abraham: Second Affidavit, November 9, 2007
On November 9, 2007, Lieutenant Colonel Stephen Abraham signed an
affidavit in response to a request by the Federal Public Defender for the District of Oregon. This was a request to provide additional information about the manner in which OARDEC operated during his assignment there, from September 11, 2004 until March 9, 2005. In this affidavit Lt. Col. Abraham also comments upon certain declarations provided by the directors of the national intelligence organizations filed in the Detainee Treatment Act litigation (see par. 3). CSHRA presents this affidavit below in searchable text. For the original affidavit, click here.
I, Stephen Abraham, hereby declare as follows:
1. I am a Lieutenant Colonel in the United States Army Reserve, having been commissioned in 1981 as an officer in the Intelligence Corps. I have served in military intelligence from 1982 to the present during periods of reserve and active duty, including mobilization in 1990 during Operation Desert Storm and twice again following September 11, 2001. My latest mobilization before my assignment to the Office for the Administrative Review of the Detention of Enemy Combatants (“OARDEC”) was as the Lead Counterterrorism Analyst for the Joint Intelligence Center, Pacific Command, from November 13, 2001 through November 12, 2002, for which I received the Defense Meritorious Service Medal. In that capacity I became very familiar with the wide variety of military intelligence resources and techniques used in the fight against terrorism.
2. In my civilian life I am an attorney with the law firm Fink and Abraham in Newport Beach, California.
3. I have been asked by the Federal Public Defender for the District of Oregon, to provide this additional information about the manner in which OARDEC operated during my assignment there, from September 11, 2004 until March 9, 2005. I have also been asked to comment upon certain declarations provided by the directors of the national intelligence organizations (“Intelligence Directors” and “Directors’ Declarations”), that I understand were filed in the Detainee Treatment Act litigation in support of Respondent’s motion to stay of production of the “Government information.”
THE OARDEC MISSION
4. When I was assigned to OARDEC in September 2004, its mission was to conduct military tribunal hearings in order to test the validity of prior summary pronouncements that the detainees at Guantanamo had been properly classified as Enemy Combatants. OARDEC rules authorized the collection of relevant information prior to the empanelling of a three-officer Combatant Status Review Tribunal that would review the evidence and make the decision. Most of the staff assigned to OARDEC were volunteer reserve forces with little or no experience dealing with intelligence or legal matters.
5. A Memorandum dated 29 July 2004, set forth the Combatant Status Review Tribunal Procedures. The procedures for requesting “Government Information.” appeared at Enclosure (“Encl.”) 1.E(3). The Memorandum defined “Government Information” as “such reasonably available information in the possession of the U.S. Government bearing on the issue of whether the detainee meets the criteria to be designated as an enemy combatant, including information generated in connection with the initial determination to hold the detainee as an enemy combatant and in any subsequent reviews of that determination, as well as any records, determinations, or reports generated in connection with such proceedings[.]”
6. I understood the 29 July 2004 Memorandum, generally, and the Enclosures thereto, including Enclosure 1, specifically, to constitute some of the guidelines by which I should perform my duties while at OARDEC.
7. The foregoing instruction authorized OARDEC to request information from numerous United States Government agencies, including those agencies with primary responsibility for gathering the intelligence. The instruction set forth no limitation relating to the classification level of material that OARDEC was capable of receiving in order to perform its primary mission.
OARDEC FAILED TO COLLECT THE INFORMATION NECESSARY TO ACHIEVE ITS MISSION
8. I realized soon after arriving at OARDEC that there were a number of problems relating to information to be used by the Tribunals. Firstly, the facilities and systems utilized by OARDEC precluded access to or use of information that OARDEC needed in order to perform its primary mission effectively. Secondly, there was no systematic method for requesting the government information relating to specific detainees. Thirdly, the individuals collecting, reviewing, and processing the information to be used by the Tribunals appeared to have little experience with intelligence products. As a result, they rarely selected the most promising sources of information and failed effectively to identify and pursue leads if any developed. While many of the detainees at Guantanamo were unknown to the intelligence agencies before their detention, given its limitations, OARDEC was unable to obtain much of what information did exist.
9. Firstly, the information directly available to OARDEC (DC) was limited to material classified SECRET and below. The information systems used by personnel of OARDEC (DC) did not permit access to TOP SECRET information. Moreover, the facility in which OARDEC (DC) personnel operated did not permit utilization of systems designed or intended to process access to TOP SECRET information and, furthermore, did not allow for the handling and/or storage of TOP SECRET information. These limitations significantly hindered OARDEC’s ability to perform its primary function.
10. The most efficient approach to information collection would have been for OARDEC to send requests for information to intelligence organizations that might have been able to guide it, ultimately, to the best sources of information. However, because OARDEC used less secure systems and facilities than those organizations, personnel at OARDEC (DC) did not have any means of communicating requests for information electronically to those organizations, including the CIA and NSA.
11. Moreover, because of the constraints on OARDEC (DC)’s systems, even if OARDEC had been granted access to more highly classified information, it could not have retained or utilized that information. This limitation precluded any possibility that such sensitive information could be incorporated into materials presented to the Tribunals.
12. Most individuals at OARDEC (DC), including the researchers, recorders, and personal representatives, were unfamiliar with, and some were unaware of, the more securely classified systems and information utilized by the intelligence organizations.
13. None of the extremely sensitive information described in the Directors’ Declarations, particularly including information relating to sensitive HUMINT sources, methods, or operations, was accessible via the systems utilized by OARDEC (DC).
Related Constraints on Team Members’ Collection of Information
14. OARDEC (DC) members’ access to information was restricted even within their limited network architecture. On the databases available to OARDEC members, access to much information was confined to particular individuals or groups, called communities of interest (COI’s). In order to access COI-restricted information, individuals either had to be members of the COI or obtain special access. Even if an OARDEC member had the appropriate clearance and access to the overall system, without a password and authorization, he or she would be denied access to COI information.
15. I observed that most of the individuals responsible for conducting research in support of the Recorder function lacked access to COI-restricted information. If there were information about a detainee in those other systems, the OARDEC researchers could not find it.
16. These researchers had little if any understanding of the nature of or even the existence of the myriad of intelligence components, whether military or otherwise. They literally did not know what they were missing.
17. Adding to the difficulties facing OARDEC was the directive that the all of the more than 500 CSRTs be completed within 120 days if possible. Even with more resources, it would not have been possible to complete these CSRTs in that limited period of time with any confidence in the reliability of the results.
Reliance on Outside Agencies
18. Another serious limitation on the ability of OARDEC to perform its mission was the fact that OARDEC was entirely dependent on indulgences from external organizations. It had no organic intelligence assets, no collection capabilities, no dissemination authority, and no direct tasking authority. OARDEC (DC) was neither in a position to demand nor expect fulfillment of its information requests, and certainly not where it offered no personnel support.
19. Even when OARDEC was able to break through the institutional barriers existing between agencies, it had no coherent process for requesting information from any particular, let alone the most appropriate, sources.
20. Upon my arrival at OARDEC, I learned that only a few individuals, including one assigned civilian, were responsible for all coordination with external organizations. The process of coordination involved two sorts of requests, standing requests issued at the beginning of a detainee tribunal cycle and ad hoc requests.
21. Standing requests consisted of broadly worded requests for any information about a detainee. As an initial matter, these general requests would go to a limited number of outside agencies.
22. These requests were directed to agencies over which OARDEC had no tasking authority. As a result, whether the agency responded to OARDEC was largely dependent on whether anyone at the agency was inclined to do so.
23. Adding to the formidable issues inherent in attempting to persuade other agencies to devote their resources to OARDEC’s effort, OARDEC’s “data calls” were sent to other organizations at the same time that the detainee was notified of his hearing. This imposed a time limit of between 30 and 40 days for the agency to respond, a condition all but assured a limited response, if any.
24. In most instances, OARDEC received either a negative response (no information available) or no response at all to its requests for information from outside agencies.
25. The second sort of OARDEC request consisted of ad hoc requests, so called “requests for information” (“RFI’s”). These were “one time” requests for specific information.
26. RFI’s were occasionally, but certainly not in every instance, initiated by a team member preparing materials for use by the Recorder. Rarely would the team member specify a specific agency, intelligence or otherwise, to which the RFI should be directed. The RFI would be sent to an RFI coordinator who would select the recipient. Until my arrival, most of these requests were sent to only a few agencies, and none were sent to the agencies that traditionally would have had responsibility for the sort of information sought.
27. The process for making ad hoc requests to outside agencies was quite arbitrary. Usually these requests were made only if some fact or other piqued the curiosity of a recorder or case researcher. There were few of these requests, and all were made within 40 days of the tribunal hearing, often less. This left insufficient time for an outside agency to respond. In my experience, the few outside agencies that received these ad hoc requests soon learned that if they did not respond within the allotted time frame, OARDEC would drop the matter or willingly accept a “negative response.”
28. During the time I was at OARDEC, although the number of information requests to outside agencies rose, there was no systematic method of making requests to the numerous agencies that one would have expected to have information about thedetainees. Partly because of the computer security problems described above, requests for information were very rarely sent to the CIA, never sent to the NSA, and never sent to the DIA. Although OARDEC faced no obstacles to communicating with the joint or unified commands, such as the Pacific Command, European Command, Central Command or Special Operations Command, few requests were directed to any of those organizations. Nor were requests sent to the U.S. Army Intelligence and Security Command until I mentioned this fairly obvious and fertile source of information.
29. Other individuals at OARDEC, including the civilian liaison officer, indicated that they preferred to deal with organizations from which they could expect some sort of response, no matter how inadequate, and rather than send requests to other entities, especially ones that had rebuffed them before or with which they had little if any experience.
30. Beyond “general data” and ad hoc information requests, OARDEC’s interaction with the outside agencies was limited to physical coordination of the various requests. This was done by two means.
31. A civilian member of OARDEC (DC) was responsible for coordinating with outside agencies. This coordination appeared limited in scope and excluded altogether at least one of the agencies from which a declaration has been provided in this case. The reason for that exclusion, as related to me by the civilian OARDEC coordinator, was that OARDEC was not cleared to receive any of that agency’s information.
32. I was involved in coordination activities between OARDEC and the outside agencies, as described in my prior declaration.
33. During the time that I was there, OARDEC relied primarily upon information provided by Joint Task Force Guantanamo, the organization tasked to deal with the detention of the detainees in the first instance. This information consisted primarily of post-detention custodial and interrogation reports.
Lack of Experience and Subject Matter Expertise
34. Apart from the foregoing obstacles to collecting information, there was a significant lack of personnel at OARDEC who were trained or experienced in collecting or utilizing intelligence information.
35. In my experience, entities that collect, process, and/or disseminate intelligence or to prepare materials derived from intelligence have a specialized staff, ranging from a few well-trained individuals to teams of hundreds. Having had experience at both ends of that spectrum, I was surprised that OARDEC lacked a staff equipped to perform its extraordinary and historic mission.
36. OARDEC’s staff should have included collection management officers, individuals trained to work with specific intelligence disciplines or specialties, and intelligence analysts. While OARDEC had assigned individuals to these tasks, most of them had little or no relevant experience, far short of the substantial experience typically found in organizations assigned far less formidable tasks than that facing OARDEC. Moreover, at least one request for the additional intelligence-trained manpower wasdenied.
37. None of my observations should have come as a surprise to any of the senior officers in OARDEC, no matter what their specific individual or collective experiences. In my years of military service, I have rarely come across any officers of notable experience, and certainly none above the rank of Major or, for the Navy, Lieutenant Commander, who had not had some experience with the intelligence process. Even if the senior officers in charge of OARDEC did not have specific experience in intelligence, most of them would have had exposure to those matters, whether in “real world” environments or through formal military schooling. The senior leadership of OARDEC should have recognized the lack of experience on the part of assigned personnel and known that a few weeks of “training” was insufficient to enable them reliably to collect and evaluate the information called for by the CSRT rules. They would have recognized that their facilities and information systems were inadequate to handle information necessary to their mission requirements. And they would have recognized that the time limitations imposed by the aggressive 120-day effort to complete all CSRT’s, coupled with the difficulty of obtaining information from external sources, all but guaranteed that they would be left to their own resources to obtain information to be presented to the Tribunals. Finally, they would have recognized that the results would fall far short of the expectations set forth in the implementing guidelines and of what was needed if the CSRTs were to produce reliable results.
Failure of Analysis
38. Information reasonably related to the process of determining whether an individual was properly classifiable as an enemy combatant (“EC”) fit into two categories: (1) information tending to prove directly that the individual engaged in activities consistent with denomination as a EC; and (2) information tending to disprove that the individual was classifiable as a EC (“exculpatory evidence”). Other information, not probative of the detainee’s EC status, was not relevant to OARDEC’s work.
39. Relevant information from both of the categories above could be divided into three different subcategories: (a) location; (b) conduct; or (c) status. Location related to whether the detainee had been in a particular place as alleged in information before the Tribunal. Conduct related to whether the detainee had engaged in particular acts as alleged in information before the Tribunal. And status related to whether anything else about the detainee tended to cast light on whether he was properly classified as an EC.
40. I observed virtually no regard on the part of team members or Recorders for the significance of these categories or subcategories of information in the investigative process.
Compiling the Government Information
41. The process of compiling material for each Tribunal would typically begin with the file received from Guantanamo. That typically contained material generated post-detention, including summaries of interrogations and incident reports (formal reports of contact with detention staff). It might include summaries or other information documenting the initial detention, including notes on the contents of items in the detainee’s possession, but this was not so in every case.
42. Seizing upon references to organizations, countries, or events in the files received from Guantanamo, the team members would conduct searches of their limited databases for additional information.
43. Of relevance to OARDEC’s mission, the Joint Detainee Information Management System (JDIMS) incorporated mostly post-detention information, including what little information was gleaned from the point of the initial capture or transfer of the detainee, interrogation reports, custody incident reports, but little specific information relating to the detainee’s history prior to his capture. Beyond JDIMS, the team members would conduct searches for material posted on the SIPERNET although, in most instances, this resulted only in the collection of finished intelligence reports that rarely related to any specific individual.
44. Most team members lacked access to additional nonpublic information, and they lacked experience working with the vast range of open source materials.
45. Rather than breaking down the question of the detainee’s status and addressing the evidence – or lack of evidence – relating to each category and subcategory, OARDEC members cast broad nets for any information, no matter how marginal, no matter how tenuous, no matter how dated, no matter how generic, no matter how dubious the source, so long as it could be connected to the detainee.
46. For the most part, the information collected lacked critical assessments of its reliability. Even where there was such an assessment, it was usually dropped from the information that was forwarded to the Recorders and Tribunals.
47. Team members would either “cut and paste” the results of their searches into electronic working documents or print them out and place them in folders.
48. Where no information was obtained about an individual, as was the case for nearly all detainees except individuals of prominence, the search would shift to more broadly based themes, such as the region from where the individual came, his ethnic group or nation of origin, or any organization denominated as being associated with terrorist activities, with which the individual was alleged to have been associated. In the latter instance, team members presumed that having an alleged association with an organization was a sufficient basis for attributing all research relating to that organization to the individual.
49. This frequently resulted in the collection of generic, generalized information on organizations and events having no direct connection to the individual.
50. Moreover, information relating to the credibility of a source was omitted, making sources appear authoritative even when they were suspect. For example, an accusation against a group would be repeated without disclosing that it originated with one of the groups’ political opponents or some government overtly hostile to it.
EXCULPATORY EVIDENCE AND THE RELEVANCE OF SENSITIVE INTELLIGENCE
Omission of Exculpatory Evidence in the Government’s Possession
51. For the reasons identified above, OARDEC did not obtain relevant information from the various intelligence agencies and consequently failed to determine whether their files held exculpatory evidence. Moreover, because OARDEC did not have a rational system for pursuing leads that might have resulted in the discovery of exculpatory evidence, it failed even to obtain all of the exculpatory information from the entities to which it did have effective access. It is likely that some of these sources had information about at least some of the detainees.
Failure to Check Detainees’ Claims
52. Moreover, a major source of highly relevant information was never explored - independent evidence from the detainee's life before his arrest. Capitalizing on self-imposed time constraints, OARDEC failed to pursue information leads offered not only by the detainees themselves but by others that would have led, with little difficulty, to additional sources of information, theretofore unexplored, most of them unclassified, that could have been used to corroborate or refute the detainee's statements. Often, that refusal to pursue information was initiated within the Tribunal Hearing process itself where request for by the detainee were deemed irrelevant or ignored altogether.
53. Detainee claims of innocence often could have been corroborated or disproved by a few simple inquiries. For example, if a detainee told interrogators that he had worked at a hospital in Afghanistan, OARDEC could have requested that an agency with regional or functional purview locate and obtain records from the hospital and interview personnel there.
54. Obtaining information in the foregoing manner would not have required the disclosure of any classified information and certainly would not have revealed sensitive intelligence sources or methods.
Refusal to Hear Outside Witnesses
55. Not only was no independent effort ever made to verify or disprove detainee claims of innocence, the detainees were unable to obtain evidence from witnesses they themselves identified. The short period of time between notice to the detainee of the tribunal proceedings and the hearings, no more than 40 days under the best of circumstances and as short as a week in some instances, and the absence of any budget for the collection of statements or testimony (whether live or recorded), precluded use of any witnesses other than detainees.
56. Beyond impractical discussions about bringing villagers to the nearest video conference facility, I was not aware of any realistic efforts at OARDEC (DC) to identify or even attempt to bring before the Tribunal witnesses or their statements.
57. In fact, OARDEC (DC) was designed to conduct Tribunals without witnesses other than the accused detainee. Upon being notified of the Tribunal, a detainee had the option of attending. If he declined, the Tribunal was immediately scheduled to be conducted in Washington. None of the Tribunals held in Washington could accommodate witnesses.
58. Although detainees were told that they could request witnesses, and some of these requests were sent by OARDEC to the State Department, no non-detainee witnesses appeared in any Tribunal.
59. For all of these reasons, OARDEC failed to fulfill its duty to present all reasonably available evidence – and not merely some classified information from a few governmental entities – bearing on the question of whether a particular detainee was properly classifiable as an Enemy Combatants.
NATIONAL SECURITY CONCERNS
Risks of Disclosure
60. The Intelligence Directors raise the concern that a search for Government Information if conducted now, might risk disclosure of highly sensitive national intelligence information, such as source or method information, to the Court and large numbers of detainee counsel. The concerns expressed by the Intelligence Directors are unrealistic in the vast majority of cases, and could effectively be addressed should they ever arise.
The Most Sensitive Information Was Not Collected
61. The most sensitive information described in the Director’s Declarations, most notably sources and methods information, was not collected for the CSRT’s. Moreover, even assuming OARDEC (DC) had gone so far as to satisfy itself by conducting an exhaustive search of the entirety of the records of the organizations, what it would have likely discerned from the exercise is that there is little information to be obtained on people that have never before been considered let alone determined to be persons of interest.
62. Government information that has, at its source, the kinds of sensitive national intelligence information discussed by the intelligence directors is not normally shared between intelligence agencies except in the rarest of circumstances. Rather, in its place, there is a recital of an assessment of the source’s reliability or credibility or, where applicable, the fact that certain aspects of the information is independently corroborated.
63. Specifically, information about collection methods and techniques, identified in paragraph 10 of General Hayden’s declaration, would not have been disclosed to anyone at OARDEC (DC) under any circumstances. Furthermore, it would not have been disclosed to any members of the Tribunals. Again, the simple reason for this is that the information described by General Hayden and that about which he expresses concern relates to the sort of specific information that is not included in finished products disseminated to intelligence consumers except in the rarest of instances not applicable to OARDEC or the CSRT’s.
64. Consistent with these facts, at no time did I observe the use by any personnel at OARDEC (DC), whether research team members, Recorders, or Tribunal members, of any information that contained information related to “clandestine intelligence activities, including counterterrorism operations, foreign intelligence information and assistance, information provided by sensitive sources, and technical collection activities.” (Declaration of General Hayden, ¶10).
65. As described above, OARDEC was unable to collect this information, even had the outside agencies been willing to provide it. Few individuals had the appropriate clearance to receive, use, or disseminate sensitive HUMINT information, and OARDEC (DC)’s information systems were not accredited for handling or processing that information.
66. Each of these factors precluded OARDEC (DC)’s receipt of the types of information described in General Hayden’s declaration.
67. Similarly, these limitations foreclosed OARDEC from obtaining the information described in the declaration of General Alexander, Director of the National Security Agency.
68. For all of the reasons described above, OARDEC was incapable of collecting or competently analyzing much of the relevant information about the detainees held by the government.
69. Most of the information OARDEC collected about the detainees consisted instead of information obtained during interrogations of other detainees.
70. At no time did OARDEC or, to my knowledge, any other government agency, make any effort to investigate whether outside evidence corroborated these statements from other detainees. Nor were detainees able to obtain evidence from outside witnesses at their CSRTs.
71. The types of extremely sensitive information discussed in the Directors’ Declarations were never collected by OARDEC. Had OARDEC even conducted the searches in the first instance, the concerns would never be realized because (1) the sort of sensitive material described by the directors would not have been probative of the issues to be addressed by the Tribunals and would not have been shared with OARDEC under any circumstances; and (2) in the few cases where the concerns might apply, there are adequate mechanisms in place to provide for in camera review of any critical information, the nature of which precludes disclosure beyond the court.
I hereby declare under penalty of perjury based on my personal knowledge that the foregoing is true and accurate.
Dated: November 9, 2007 (Signed)
Get original here