COMMENTS BY

THE OFFICE OF THE UNDER SECRETARY OF DEFENSE

ONA

DRAFT OF A PROPOSED REPORT

BY THE DOD OFFICE OF INSPECTOR GENERAL

PROJECT NO. D2006DINTO1-0077.000

REVIEW OF PRE-IRAQI WAR ACTIVITIES

OF THE OFFICE OF

THE UNDER SECRETARY OF DEFENSE FOR POLICY (U)

January 16, 2007

OUSD(P) COMMENTS ON

DRAFT OF A PROPOSED REPORT

BY THE DOD OFFICE OF INSPECTOR GENERAL

REVIEW OF PRE-IRAQI WAR ACTIVITIES OF THE OFFICE OF THE

UNDER SECRETARY OF DEFENSE FOR POLICY (U)

PROJECT NO. D2006DINTO1-0077.000

January 16, 2007

TABLE OF CONTENTS

I. SLTMMARY OF KEY ERRORS IN THE DRAFT REPORT (U) 4

II. THE POLITICAL BACKGROUND OF THIS MATTER (U) 7

III. OUSD(P) SUPPORT TO THE OIG REVIEW (U) 11

IV. FACTS (U) 11

A. THREE SEPARATE ACTIVITIES RELATING TO THE WORK UNDER REVIEW (U) 12

1. The PCTEG (U) 13

2. The DIA Analyst Detailed to the Policy Support Office (U) 16

3. The Deputy Secretary ‘s Tasking to Brief the Secretary of Defense (U) 18

B. DRAFT BRIEFING TO THE SECRETARY OF DEFENSE (U) 22

C. THE SECRETARY OF DEFENSE’S DIRECTION TO BRIEF DCI, DRAFT BRIEFING TO DCI,

CIA MEETING (U) 25

D. DEPUTY NATIONAL SECURITY ADVISOR’S REQUEST, DSD’S DIRECTION, DRAFT

BRIEFING TO DEPUTY NATIONAL SECURITY ADVISOR (U) 27

E. DCI’S CONGRESSIONAL STATEMENTS ON IRAQ AND AL-QAIDA (U) 28

V. DISCUSSION (U) 31

A. WHY ARE LAWFUL AND AUTHORIZED ACTIVITIES NEVERTHELESS CALLED

“INAPPROPRIATE”? (U) 31

B. DIA’S DI POLICY N0S. 004 AND 005 Do NOT APPLY TO NON-IC OFFICES DIRECTED

BY SENIOR DOD LEADERS TO CRITIQUE INTELLIGENCE COMMUNITY WORK (U) 33

1. The Internal DIA Policies Do Not Apply to DIA Members While Detailed to

Policy Positions Outside DIA ‘s Chain of Command (U) 34

2. The Internal DIA Policies Contain No Procedure for an IC Customer to Obtain

an Alternative IC Judgment, Which in any Case is not What the DSD Sought Here

(U) 34

3. The Internal DIA Policies Were Not Coordinated or Published as Would Have

Been Required ~f Intended to Apply Outside DIA (U) 35

C. “INTELLIGENCE ACTIVITIES” CONSTITUTE A PROCESS USING ALL KEY ELEMENTS

OF INTELLIGENCE WORK BY INTELLIGENCE AGENCIES (U) 35

D. ALTERNATIVE OR CRITICAL ASSESSMENTS OF IC INFORMATION AND IC JUDGMENTS

BY NON-IC OFFICES ARE NOT “INTELLIGENCE ACTIVITIES” (U) 37

E. OUSD(P) DID NOT PRODUCE OR DISSEMINATE “INTELLIGENCE ASSESSMENTS” OR

“INTELLIGENCE ANALYSIS” (U) 38

F. THE RELEVANT ORDERS AND DIRECTIVES DESCRIBE INTELLIGENCE ROLES AND

ACTIVITIES, THEY Do NOT PROSCRIBE POLICY ACTIVITIES (U) 40

VI. OUSD(P) NONCONCURRENCE (U) 43

A. WITH THE FINDINGS OF THE DRAFt REPORT (U) 43

B. WITH THE RECOMMENDATIONS OF THE DRAFT REPORT (U) 44

VII. CONCLUSION (U) 45

APPENDIX A: DI POLICY NOS. 004 AND 005 47

APPENDIX B: COMMENTS ON OIG’S ANSWERS TO SENATOR LEVIN 50

OUSD(P) COMMENTS ON

DRAFT OF A PROPOSED REPORT

BY THE DOD OFFICE OF INSPECTOR GENERAL

REVIEW OF PRE-IRAQI WAR ACTIVITIES OF THE OFFICE OF THE

UNDER SECRETARY OF DEFENSE FOR POLICY (U)

PROJECT NO. D2006DINTO1-0077.000

January 16, 2007

The Committee believes that IC analysts should expect difficult and repeated questions regarding threat information. .. . The Committee found that this process — the policymakers probing questions — actually improved the JCIA ‘sJ products. ... While analysts cannot dismiss a threat because at first glance it seems unreasonable or it cannot be corroborated by other credible reporting, policymakers have the ultimate responsibility for making decisions based on this same fragmentary, inconclusive reporting.’

(U) The Office of the Under Secretary of Defense for Policy (OUSD(P)) offers the

following comments on a December 20, 2006 Draft of a Proposed Report (the “Draft

Report”) by the Department of Defense Office of Inspector General (“OIG”) in Project

No. D2006DINTO1-0077.000, “Review of Pre-Iraqi War Activities of the Office of the

Under Secretary of Defense for Policy (U)” (the “Project”).

(U) Throughout these comments we observe that the work on which this Project concentrates, and in particular the specific activities that the Draft Report characterizes as “inappropriate,” were authorized and directed to be done by the Deputy Secretary or the Secretary of Defense. For the purpose of these comments, references to “work” or “activities” “authorized” and “directed” by the Secretary, the Deputy Secretary, “the most senior leaders” of DoD, or “senior DoD leaders” specifically mean the following:

(U) The Deputy Secretary of Defense (“Deputy” or “DSD”) directed his Special Assistant in his front office and two staff members in OUSD(P) to take a fresh, critical look at Intelligence Community (“IC”) reporting on contacts between Iraq and al-Qaida. In working on the Deputy’s tasking, one of the OUSD(P) staffers prepared an internal memo containing two commentary paragraphs followed by a list summarizing IC reports on contacts between Iraq and al-Qaida. The staffers wrote up the critique requested by DSD in the form of a draft briefing that discussed IC reports on Iraq-al-Qaida contacts and how these reported contacts might be viewed absent an a priori assumption that secular Baathists and Islamic extremists would never cooperate. The Deputy Secretary

(U) Report of the Select Committee on Intelligence on the US. Intelligence Community ~ Prewar Intelligence Assessments on Iraq (U) (9 July 2004), pp. 34, 35 (unanimous report, unclassified version) (“SSCI Report”).

directed that the draft briefing be given to the Secretary of Defense. After receiving it, the Secretary directed that it be shared with the DCI. The Deputy Secretary’s office also directed that the draft briefing be given to the Deputy National Security Advisor when the latter requested it.

I. SUMMARY OF KEY ERRORS IN THE DRAFT REPORT (U)

• (U) The title of the Draft Report is inaccurate. The work on which the Draft Report focuses was not “OUSD(P)”activity. It was in fact a response to tasking by the Deputy Secretary of Defense, who in July 2002 directed his Special Assistant in his front office and two staff members in OUSD(P) to critique IC reporting on contacts between Iraq and al-Qaida. The result was a draft briefing on how those contacts might be viewed if one did not assume a priori that secular Baathists and Islamic extremists would never cooperate. The Deputy Secretary directed that the draft briefing be given the Secretary of Defense. After receiving it, the Secretary directed that it be shared with the DCI. When the Deputy National Security Advisor requested the draft briefing, the Deputy Secretary’s office directed that it be given to him.

• (U) The work reviewed was not an “OUSD(P)” activity, assessment, view, position or initiative, despite the Draft Report’s repeated assertions to the contrary. The Under Secretary of Defense for Policy (USDP) never approved, adopted or advocated the draft briefing or any of the work leading to it as an “OUSD(P)” view or assessment. Each version of the briefing was marked “draft” or “draft working papers” and was never presented as anything other than that.

• (U) The Draft Report correctly finds that these activities were lawful and authorized. It correctly states (page 34) that “the Secretary [of Defense] owns the DoD Directives governing (among others) Intelligence and Policy, and as long as Executive Orders or other legal statutes are not violated, he has the latitude to interchange roles and responsibilities.” But in contradiction of these same findings, the Draft Report incorrectly calls the activities “inappropriate,” because they supposedly amounted to “dissemination” to senior decision-makers of “alternative intelligence assessments” “inconsistent” with the “consensus” of the IC.

• (U) If the OIG believes that it was inappropriate for the Deputy Secretary of Defense to have non-IC OSD staff members critique IC work on a significant subject of national security, inappropriate for the Secretary of Defense to share the OSD work with the DCI, and inappropriate for the Deputy Secretary to share the work with the Deputy National Security Advisor when requested by the latter, the OIG should say so directly instead of finding fault with subordinate OSD offices and staff members who did as the Secretary or Deputy Secretary instructed.

• (U) The entire argument in the Draft Report rests on the definition of “Intelligence Activities” and the meaning of “intelligence assessments.” The Report’s interpretation of the definition of “Intelligence Activities” found in the relevant DoD directive is wrong. By its definition, that term on its face applies only to intelligence agencies, not to policy offices.

• (U) Because OUSD(P) routinely and properly acquires, assesses and distributes “information relating to the capabilities, intentions, and activities of foreign powers,”² stretching the definition of “Intelligence Activities” to include policy offices would lead to the absurd result of mischaracterizing most work done in OUSD(P) as “Intelligence Activities.”

• (U) The Report does not define the term “intelligence assessments” but erroneously asserts that a critique by non-IC staffers of IC assessments was itself an “inappropriate” “intelligence assessment.” There are no facts in the Draft Report, or otherwise, supporting the assertion that this work was presented as “intelligence assessments.”

• (U) There are likewise no facts suggesting that the “senior decision-makers” who were briefed on this work, specifically, the Secretary of Defense, the Deputy Secretary of Defense, the DCI, the Deputy National Security Advisor, and the Vice President’s Chief of Staff~, mistook this work to be “intelligence assessments.”

• (U) The Report fails to make clear that the Office of Special Plans (OSP), the Policy Counter Terrorism Evaluation Group (PCTEG), and the Policy Support Office did not perform and had no responsibility for any of the work reviewed in this Project. This failure is especially egregious in light of press reports and political criticism that continue to assert the contrary.

• (U) The Draft Report labels the work products at issue as “inappropriate” (page 4) because they allegedly “did not clearly show the variance with the consensus of the Intelligence Community” and “were, in some cases, shown as intelligence products.” But the senior decision-makers briefed on this work (one of whom was the DCI himself) did not need to be told that it varied in some respects from IC analysis; that was inescapably obvious. There are no facts to suggest that any of them drew any conclusions or made any decisions whatsoever solely on the basis of the draft briefing, without taking IC views into account. There are no facts supporting the claim that some work products “were, in some cases, shown as intelligence products.”

2 (U) Part of the definition of “foreign intelligence,” which in turn is part of the definition of “Intelligence Activities.” See DoD Directive No. 5240.1, DoD Intelligence Activities,” 25 April 1988, Sections 3.1 and 3.2.

• (U) OUSD(P) did not impede or undercut any responsibilities of the Intelligence Community, contrary to suggestions in the Draft Report. The IC was fully aware of the work under review and commented on it several times, as the Draft Report itself reveals. Further, the DCI was personally briefed on the work at the Secretary of Defense’s direction.

• (U) OUSD(P) did not bypass any applicable DIA procedures, contrary to assertions in the Draft Report. The DIA’s DI Policy Nos. 004 and 005, cited by the Draft Report, are internal DIA guidelines that only apply to DIA analysts, working as such, who wish to produce alternative analyses or alternative judgments within DIA’s chain of command. These guidelines are irrelevant to customer offices of the IC — the consumers of intelligence -- that wish to suggest an alternative way of viewing information and analyses already provided by the IC. Nor do these guidelines provide any mechanism for DIA customers to request an alternative judgment by the IC, which in any case is manifestly not what the Deputy Secretary desired when he directed the work under review to be done.

• (U) While some of the work reviewed in this Project did characterize the Iraq-alQaida relationship as “cooperative,” that characterization did not contradict IC judgments on the subject at the time. To the contrary, the reference in the draft briefing to a “cooperative” Iraq-al-Qaida relationship was consistent with the DCI’s own statements to Congress in 2002 and 2003. He said then that “we have solid reporting of senior level contacts between Iraq and al-Qaida going back a decade,” “credible information indicates that Iraq and al-Qaida have discussed safe haven and reciprocal non-aggression,” “we have solid evidence of the presence in Iraq of alQaida members,” “the reporting also stated that Iraq has provided training to al-Qaida members in the areas of poisons and gases and making conventional bombs,” etc. The Draft Report ignores these DCI statements.

• (U) The Draft Report erroneously faults OUSD(P) for failin~ to provide “the most accurate analysis of intelligence” to senior decision-makers. That responsibility rests with the IC, not OUSD(P). More importantly, senior decision-makers already had the IC’s reports and assessments on Iraq and al-Qaida and thus already had “the most accurate analysis of intelligence” -- if one accepts, as the Draft Report seems to do, that the IC’s assessments are the “most accurate.”⁴

~ (U) This criticism is symptomatic of the peculiar and sometimes contradictory logic of the Draft Report, for the Draft Report also holds that OUSD(P) should not provide any intelligence analyses at all.

~ (U) The Draft Report purports to make judgments about the nature of the Iraq-al-Qaida relationship, but these judgments appear to be based on certain CIA and DIA analytical papers --not on any contemporaneous ME or other authoritative consensus by the IC as a whole -- and without reference to the DCI’s own statements on the subject. There is no evidence in the Draft Report that the OIG

• (U) The Draft Report recommends (page 14) that, if OUSD(P) disagrees with an IC consensus, OUSD(P) should “clearly articulate in policy products the Intelligence Community consensus and the basis for disagreement or variance from the Intelligence Community consensus.” Such a requirement would inappropriately constrain policy work by requiring policy offices to vet every policy recommendation or analysis with the IC in order to determine whether or not it disagreed or varied with an IC “consensus.” It would also burden policy offices with a requirement to articulate the IC “consensus” when the IC itself should do so.

• (U) Bipartisan reports and studies by various commissions and congressional committees since the 9/11 attacks have stressed the need for vigorous debate, hard questions and alternative thinking of the sort that motivated the work reviewed in this Project. The conclusions and recommendation in the Draft Report reflect a disturbing departure from the lessons of all these reports and studies. By faulting a critical assessment in OSD of IC work on contacts between Iraq and al-Qaida, the Draft Report would inhibit the vigorous debate and hard questioning that most observers recognize as essential. The Draft Report’s conclusions, if sustained, would have a dampening effect on future initiatives challenging intelligence assessments. The facts do not justify such conclusions.

II. THE POLITICAL BACKGROUND OF THIS MATTER (U)

(U) The activities reviewed in this Project, unfortunately, have been the object of bitter political debate and inaccurate press reporting for over three years. Given the partisan nature of the matter, it was particularly important that the OIG’s independent review adhere to the strictest standards of factual accuracy, rigorous analysis, and clarity of expression. Unfortunately, the Draft Report does not meet those standards.

(U) Apart from numerous factual inaccuracies, omissions and mischaracterizations identified throughout these comments, the Draft Report suffers from a basic analytical flaw in attempting to paint the work under review as “inappropriate” even though no laws were broken, no DoD directives were violated, and no applicable policies were disregarded. The Draft Report concedes that the activities reviewed were lawful. It concedes that the activities were authorized — indeed requested — by the Deputy Secretary and Secretary of Defense. In perhaps its most trenchant observation, the Draft Report correctly states (page 34) that “the Secretary owns the DoD Directives governing (among others) Intelligence and Policy, and as long as Executive Orders or other legal statutes are not violated, he has the latitude to interchange roles and responsibilities” (emphasis added).

undertook any rigorous, independent review of the underlying intelligence on the issue of contacts between fraq and al-Qaida.

(U) That observation goes to the heart of the present matter. It shows that the activities in question were clearly appropriate. No statutes or executive orders were violated. The Secretary, and by extension the Deputy, unequivocally had the latitude to obtain an alternative, critical assessment of IC work on Iraq and al-Qaida from non-IC OSD staff members rather than from the DIA or the Assistant Secretary of Defense for C3I, without vetting such critique through any Intelligence Community process. The Secretary had the latitude to direct the authors of such critique to share it with the DCI. The Deputy Secretary had the latitude to direct the authors of such critique to share it with the Vice President’s Chief of Staff and the Deputy National Security Advisor when the latter so requested. This should have put an end to any question of appropriateness.

(U) The OIG is empowered and competent to determine whether the activities were lawful and authorized. But we question whether it is “appropriate” for the OIG to venture into the realm of opinion about whether the activities were appropriate, in the absence of any applicable standards, regulations, directives, etc. This is especially true where, as here, the OIG has found the activities in question were lawful and authorized, and has conceded that the Secretary and Deputy have the “latitude to interchange roles and responsibilities” in overseeing DoD.

(U) We respectfully observe that the OIG’s opinion on the subjective question of “appropriateness” in these circumstances is not entitled to any particular deference. The OIG does not have special expertise on this issue, which is fraught with policy and political dimensions. Given the politically charged atmosphere infecting this entire matter, it is especially objectionable for the OIG to obscure and minimize the fact that the Secretary and Deputy Secretary directed the activities in question be done, to mischaracterize the work as “OUSD(P)” activities, and to find something “inappropriate” in the fact that subordinate offices and staffers did as the Secretary and Deputy directed.

(U) Moreover, the Draft Report employs a demonstrably incorrect reading of “Intelligence Activities” to portray the work reviewed as “alternative intelligence assessments,” “Intelligence Production” and the like, when in fact it was not. This mischaracterization is particularly egregious in light of the persistently false press reports and political accusations claiming that the Deputy Secretary, or OUSD(P), or others in the Defense Department distorted intelligence in order to argue that Iraq had a direct role in the 9/11 attacks, or that Iraq and al-Qaida had a stronger relationship than shown by facts known at the time, in order to propel the United States to war on false pretenses.

(U) Before the OIG ever took up this matter, it had been the subject of an exhaustive investigation that the Senate Select Committee on Intelligence (SSCI) began in July 2003, as well as a “minority inquiry” begun by Senator Carl Levin in June 2003.

(U) In July 2004, the Committee issued a unanimous report on “Phase I” of its investigation. That report concluded inter alia that policymakers at no time pressured the

IC to change its conclusions on Iraq’s links to terrorism, and that the work of OSD staffers reviewed here did not result in any changes to the analytical judgments in IC work on Iraqi support for terrorism.⁵The Committee deferred to a second phase of its investigation an evaluation of whether the work products now under OIG review were “objective, reasonable, and accurate.”⁶Because of divisions along partisan lines within the SSCI, its members have not to date been able to agree on what conclusions to reach in its “Phase II” report.

(U) SSCI Chairman Pat Roberts referred this matter to OIG only because these partisan divisions prevented the SSCI from reaching agreement on what to say about the activities reviewed in this Project. By the time he made the referral to OIG, the issue had been transformed from whether the work in question was “objective, reasonable and accurate” to whether it was “unauthorized, unlawful or inappropriate” -- even though the SSCI haduncovered no information to support such a characterization.

(U) In his September 9, 2005 letter requesting an independent review by OIG, Chairman Roberts wrote that “the Committee is concerned about persistent and, to date, unsubstantiated allegations that there was something unlawful or improper about the activities of the Office of Special Plans within the office of the Under Secretary of Defense for Policy during the period prior to the initiation of Operation Iraqi Freedom.” He added that he had “not discovered any credible evidence of unlawful or improper activity, yet the allegations persist.” He nevertheless asked the OIG to review “whether the personnel assigned to the Office of Special Plans, at any time, conducted unauthorized, unlawful or inappropriate intelligence activities.”

(U) On September 22, 2005, Senator Carl Levin wrote in his capacity as Ranking Member of the Senate Armed Services Committee (SASC), asking the OIG to expand the scope of the review requested by Chairman Roberts. Specifically, Senator Levin requested that “you include all elements of the Office of the Under Secretary of Defense for Policy, including the Policy Counter Terrorism Evaluation Group (PCTEG) and the Policy Support office.” He posed a number of questions for the OIG to answer.⁷

(U) In fact Senator Levin had already published his own conclusions on this matter nearly a year before the OIG took up its review. See “Report of an Inquiry Into the Alternative Analysis of the Issue of an Iraq-al Qaeda Relationship” (October 21, 2004), containing numerous incorrect allegations of improper conduct within OUSD(P).

(U) SSCI Report, p. 363. ⁶(U) SSCI Report, p. 312.

‘ (U) At Appendix B attached to these comments, we address in detail the Draft Report’s answers to

Senator Levin’s questions.

That report was part of the “minority inquiry” that Senator Levin has been pursuing into the subject matter of this Project since June 2003, without the endorsement of the SASC, the SSCI, or any other congressional committee as of early January 2007. The Draft Report (page 1) comments that Senator Levin’s report “challenged some of the conclusions” in the SSCI’s report of July 2004 but fails to note that Senator Levin himself, as a SSCI member, concurred in that same SSCI report and that the SSCI report was unanimous.

(U) It bears emphasis that the same set of facts and documents have been available to the SSCI and to Senator Levin throughout this process.

(U) More recently, on December 8, 2006, Representative Cynthia McKinney introduced articles of impeachment against the President of the United States, the first article of which makes the false assertion that the President and the Secretary of Defense created the OSP “to override existing intelligence reports by providing unreliable evidence that supported the claim that Iraq’s alleged weapons of mass destruction posed an imminent threat to the United States of America.”⁸

(U) Meanwhile, uninformed and inaccurate press reports have persisted, generally on the theme that the Office of Special Plans allegedly conducted a rogue intelligence operation before the Iraq war and fed incorrect or exaggerated intelligence information to senior policy makers in the Executive Branch, bypassing the Intelligence Community and contributing to an ill-informed decision to go to war in Iraq. These stories have been repeated so many times that they are now taken as established truth by some members of Congress and many commentators.

(U) Indeed, even the Draft Report to some extent seems to fall prey to the hypnotic effect of these constantly repeated falsehoods. Instead of setting the record straight clearly and directly, the Draft Report relegates to a footnote (at page ii, repeated at page 1) the peculiar comment that:

“The term Office of Special Plans has become generic terminology for the activities of the OUSD(P), including the Policy Counter Terrorism Evaluation Group and Policy Support Office. The actual Office of Special Plans had no responsibility for and did not perform any of the activities examined in this review.”

(U) As the facts detailed below demonstrate, neither the OSP, the PCTEG, nor the Policy Support Office had any responsibility for the activities reviewed, and none of these units as such performed any of those activities. The Draft Report should say so forthrightly.

8 (U) H. Res. 1106, 109th Cong., 2” Sess. (8 December 2006).

(U) The Draft Report should also say prominently and forthrightly that the most senior leaders of DoD directed these activities to be done by non-IC OSD staff members, not all of whom were even assigned to OUSD(P), rather than repeatedly mischaracterizing these actions as “OUSD(P)” activities.

(U) These and other deficiencies of the Draft Report, discussed in these comments, demonstrate that the OIG should reconsider its excursion into the policy and political issue of whether the lawful and authorized activities under review were “appropriate.”

III. OUSD(P) SUPPORT TO THE OIG REVIEW (U)

(U) To assist the OIG in its review, this office provided copies of the thousands of pages of documents that we had already provided to the SSCI and to Senator Levin. We also provided various additional materials that the OIG requested. In addition, we arranged for the OIG to review certain documents that DoD had earlier declined to provide the Congress. We offered OIG the opportunity to review some ten file boxes containing all the documents we had collected in the course of our initial search in response to the SSCI’s and Senator Levin’s document requests, including documents that on review we had determined to be unresponsive and thus did not provide to Congress. We also provided all witnesses that we were in a position to produce for interviews requested by the OIG and suggested various additional individuals as possible witnesses.

IV. FACTS (U)

(U) Because of the need for a clear, complete and accurate account of the relevant facts, we provide a detailed statement of facts below. Throughout the factual narrative, we undertake to highlight the more significant factual errors in the Draft Report.

(U) A discussion section, examining the authorities and analysis set out in the Draft Report, follows the statement of facts.

(U) The Draft Report does not explain the origin or context of the work under review. By persistently mischaracterizing this work as “OUSD(P)” activities, the Draft Report conveys an incorrect impression that this work was an “OUSD(P)” initiative constituting an “inappropriate” intrusion into “intelligence functions that are the responsibility of Defense Intelligence” (page 14). The Draft Report mentions that “some of the actions were performed in response to inquiries from the Deputy Secretary of Defense and direction from the Secretary of Defense” (page 13), leaving the incorrect impression that such actions were somehow incidental to other (unspecified) actions attributable solely to the “OUSD(P).”

(U) In fact, all (not some) of the work characterized by the Draft Report as

“inappropriate,” specifically, three versions of a draft briefing on links between Iraq and

al-Qaida and an internal staff memo done in preparation for the briefing, was in response to requests and taskings by either the Deputy Secretary or the Secretary of Defense. The Deputy Secretary directed that the draft briefing be prepared for the Secretary. After the Secretary received the draft briefing, he directed that it be shared with the DCI. When the Deputy National Security Advisor requested the draft briefing, the Deputy Secretary’s office directed that it be given to him. Three OSD staff members had the primary responsibility to do this work. Two happened to be DIA analysts detailed to OUSD(P) and the third worked directly for the Deputy Secretary as his Special Assistant.

(U) How and why these particular three individuals became involved in this work were as follows:

A. Three Separate Activities Relating to the Work Under Review (U)

(U) There were three, initially separate, activities within the Office of the Secretary of Defense (0 SD) that relate to the work under review in this Project. Some of the individuals involved in these three activities, and some strands of their work, eventually came together under the direction and oversight of the Deputy Secretary of Defense (DSD), who tasked certain work discussed below. That work, and certain resulting draft documents (critiquing IC work on the Iraq-al-Qaida relationship), are the actions that the Draft Report mischaracterizes as “alternative intelligence assessments” and “Intelligence Activities.”

(U) In its “Background” section the Draft Report discusses the OSP (page 3) but fails to make clear in the text that the OSP had nothing to do with any of the activities under review. None of this work or the resulting documents was done by, for, or under the direction of the OSP. The work reviewed in this Project was substantially completed before the OSP even came into defacto existence in mid-August 2002. (The Draft Report states that OSP was created in October 2002; it was in that month that certain formalities were implemented.) The Draft Report also errs in stating that the OSP was “disbanded” in July 2003. In fact it was merely renamed as the Office of Northern Gulf Affairs, remaining in NESA as before, and its personnel continued to perform their policy functions regarding that region.

(U) Likewise, none of this work or the resulting documents was done by, for, or under the direction of the PCTEG or the Policy Support Office as such.

(U) Nor did the Under Secretary of Defense for Policy ever approve or adopt any of the draft opinions or conclusions in any of the resulting documents as OUSD(P) positions, views or conclusions.

1. The PCTEG (U)

(U) The first activity relevant here was an ad hoc group, formed by the Under Secretary of Defense for Policy (USDP) shortly after the 9/11 attacks. The mission of that group was to review all available information about a number of international terrorist organizations with a basic focus on the question: What does it mean to be at war with a terrorist network?⁹The Draft Report erroneously states that this group was formed “to conduct an independent analysis of the al-Qaida terrorist network” (page 2). In fact, the group’s work was not limited to al-Qaida but addressed more generally various major terrorist groups and their relations with their state sponsors. This group commenced work in approximately October 2001 with two members: a consultant, and a detailee from the Defense Threat Reduction Agency. The group requested and received relevant intelligence information from the Intelligence Community and did preliminary work on the subject assigned. Both members, however, left for other duties towards the end of 2001 and the beginning of 2002. Neither of them ever worked in or took direction from the OSP or the Policy Support Office.

(U) In January 2002 the USDP decided to continue the project in a more formal way, by naming the project the “Policy Counter Terrorism Evaluation Group” (PCTEG) and formally requesting detailees from DIA. The memo approving creation of the PCTEG described its task as follows:

• (U) Study al-Qaida’s worldwide organization including its suppliers, its relations with States and with other terrorist organizations (and their suppliers).

• (U) Identify “chokepoints” of cooperation and coordination.

• (U) Identify vulnerabilities.

• (U) Recommend strategies to render the terrorist networks ineffective.’⁰

(U) Also, as early as January 2002, the Deputy Secretary among others was raising questions about possible links between Iraq and the al-Qaida terrorist network.” In addition to the information and analyses he regularly received through established

~ (U) Statement of Douglas J. Feith, Under Secretary of Defense for Policy, on the Policy Counter

Terrorism Evaluation Group, before the Select Committee on Intelligence, U.S. Senate, 10 July 2003,

USDP and Senator Levin Correspondence, November 03-July 05, Tab 14.

‘° (U) Memo from ASD (ISA) to USDP (31 January 2002), USDP Congressional Correspondence November 02-February 04, Tab 18.

¹¹(U) Memo from DSD to USDP (22 January 2002), reproduced as Appendix E to the Draft Report.

intelligence channels, the Deputy also asked for input from OUSD(P), including in a memo to the USDP on January 22, 2002.12 He received a reply from the Assistant Secretary of Defense for International Security Affairs on January 24, 2002, summarizing information suggesting “few direct links” and other information “suggesting more robust indirect links.”³There was nothing unusual or improper about this. How to assess the information provided by the IC and what, if any, decisions to make or conclusions to draw from it are central responsibilities of the Deputy and other senior policy officials of the Defense Department. It was not remarkable that the Deputy consulted OSD policy offices as well as the IC on possible links between Iraq and al-Qaida.

(U) In February 2002 USDP requested the Director of DIA to provide three detailees to the PCTEG.¹⁴In response, DIA provided two of the three individuals requested, both reserve Naval intelligence officers then assigned to the J-2. Contrary to the Draft Report (page 2), these officers were not detailed to OUSD(P) in October 2001; rather, they were detailed in February 2002, as replacements for the two original members of the PCTEG who were gone by the time the two DIA detailees arrived. One of these DIA detailees departed in April 2002, leaving only one member of the “group,” who continued to work as the sole member of the PCTEG until he was demobilized from Naval reserve duty in January 2003.15

(U) The PCTEG member who departed in April 2002 never worked in or took direction from the OSP or the Policy Support Office, nor did the sole remaining PCTEG member at any time relevant here.’⁶

(U) As originally conceived, the PCTEG was to function under the joint chairmanship of the Principal Deputy Assistant Secretary of Defense for SO/LIC and the Deputy Assistant Secretary of Defense for NESA’⁷(not by the ASD (ISA) and ASD (SO/LIC) as the Draft Report incorrectly implies at page 2). But the group never had

12 (U) Ibid.

13 (U) Memo from ASD(ISA) to DSD (24 January 2002), reproduced as Appendix F to the Draft Report.

14 (U) Memo for Director, Defense Intelligence Agency (2 February 2002), ibid.

15 (U) Roster of PCTEG and Special Plans/Northern Gulf, USDP Congressional Correspondence November 02-February 04, Tab 1 6A.

16 (U) After being demobilized from Naval reserve duty in 2003, the former single remaining PCTEG

member did return to OUSD(P) and worked as a civilian in OSP for a time, but that was after the work

relevant to this Project had been completed.

17 (U) Memo from ASD (ISA) to USDP (31 January 2002), USDP Congressional Correspondence

November 02-February 04, Tab 18.

more than two members and soon dwindled to one; thus it never attained the degree of operational formality implied by this nominal joint chairmanship. When the DSD began to take a more active role on the specific issue of the relationship between Iraq and alQaida, as discussed below, the single remaining member of the PCTEG participated with others in replying to DSD taskings and at times responded directly to the DSD in that regard. At no time did the PCTEG report to or take direction from the OSP or the Policy Support Office.

(U) The PCTEG produced a 154-page draft briefing entitled “Understanding the Strategic Threat of Terror Networks and their Sponsors,” which was revised and updated periodically. Consistent with the mission of the PCTEG, this briefing examined the methods and operations of various terrorist organizations (including but not limited to al-Qaida), the nature of their ties with their state sponsors, and various policy considerations on dealing with the threat posed by these groups.

(U) This briefing was the sole substantive work product by the PCTEG as such. The briefing was separate from the work, addressed in the Draft Report, on the specific issue of the Iraq-al-Qaida relationship.’⁹The PCTEG briefing was an internal Policy staff-level product that was never presented outside the Policy organization and never approved by senior policy makers as an official OUSD(P) position, so far as any facts known to us are concerned, and the Draft Report does not contend otherwise. Indeed, the Draft Report only briefly mentions but does not discuss this solitary PCTEG product.

(U) The Draft Report mischaracterizes events in stating (page 3) that the one remaining PCTEG member created a briefing in the summer of 2002 on links between Iraq and al-Qaida “with the assistance of a member of OUSD(P)’s Policy Support Office and a Special Assistant to the Deputy Secretary of Defense.” Here and throughout, the Draft Report ignores or downplays the central fact that the Deputy Secretary of Defense directed the work to be done, as discussed more fully below. He gave the assignment initially to his Special Assistant, not to the sole PCTEG member or the Policy Support Office staffer or anyone else in OUSD(P). The latter two individuals did participate in responding to the Deputy Secretary because of the circumstances explained in these comments. But it is a gross distortion to suggest, as the Draft Report does, that the sole PCTEG member originated this effort or that it was an OUSD(P) activity.

(U) The Draft Report also mischaracterizes events in stating (at page 3) that “OUSD(P) dissolved the PCTEG shortly” after the draft briefing was given to the Deputy

¹⁸(U)Ibid,Tab 15.

19 (U) Although this PCTEG briefing was separate from the work on the specific issue of the fraq-alQaida relationship done elsewhere in OSD, it obviously overlapped to a degree and eventually led to the one remaining PCTEG member’s being included in the work on that single issue.

National Security Advisor and the Vice President’s Chief of Staff in September 2002.

There was no formal action dissolving the PCTEG; rather, the “group” withered away

when its sole remaining member was demobilized from Naval reserve duty in 2003.

2. The DIA Analyst Detailed to the Policy Support Office (U)

(U) The second activity relating to the work under review was begun by a career DIA analyst whom DIA had detailed, in January 2002, to the former Policy Support Office of the Deputy Under Secretary of Defense (Policy Support) in OUSD(P). At no time did this detailee work in or take direction from the OSP or the PCTEG.

(U) DIA detailed this analyst in response to the USDP’s by-name request. Although the Draft Report states (page 2) that the Policy Support Office requested this DIA analyst due “to the voluminous amounts of intelligence the office was receiving but was unable to assess,” the quoted phrase does not appear in USDP’s request. This analyst was a 25-year intelligence veteran who, at the time of USDP’s request, was assigned to the Interagency Damage Assessment Team for the Robert Hanssen case. This analyst had had previous experience, inter alia, providing intelligence support to policy levels as well as experience in Foreign Denial and Deception analysis that the USDP needed to support certain intelligence-related duties then assigned to the Policy Support Office.²°

( ) This analyst was tasked in the Policy Support Office to provide policy support for special access programs and to carry out other duties requiring a review of various intelligence products. Sometime in early 2002, in the course of her work, she came across a finished 1998 CIA report on Iraq’s

The report mentioned that Usama Bin Laden had requested and received certain training from an Iraqi service. On her own initiative, she requested and received through CIA channels the underlying information on which the item was based, consisting of two Memo Dissems, and subsequently obtained additional CIA reports from DIA and CIA on the issue of Iraq and al-Qaida.²’

(U) As this was the only reporting that this analyst had seen on Bin Laden in this connection, and because she considered it important data for a discussion on Iraqi intelligence and all-Qaida, she wrote a one-page “assessment” (in her words) of the IC reporting and gave it to the DUSD (PS), ASD (ISA), USDP, and DSD.²²

20 (U) Memo from IJSDP to Director, DIA (23 November 2001), USDP Congressional Correspondence

November 02-February 04, Tab 17; Memorandum for the Record (30 October 2002), ibid.

²¹(U) Memorandum for the Record (17 April 2002), Memorandum for the Record (30 October 2002),

USDP Congressional Correspondence November 02-February 04 Tabs 17 and 18.

22 (U) Memorandum for the Record (17 April 2002), ibid at Tab 17.

(U) Again on her own initiative, in early spring 2002 the analyst met with the chief of DIA’s Joint Intelligence Task Force for Combating Terrorism (JITF) and gave him a copy of the reference in the finished CIA report, the two underlying reporting documents, and her one-pager. (This one-pager should not be confused, as the Draft Report seems to do, with a later, July 25, 2002 internal memo that this analyst wrote in preparation for the August 2002 briefing to the Secretary of Defense, discussed below.²³) She recommended that the JITF publish the IC reporting data “so that it would be available to the entire IC because reports published previously did not contain this important data” and that, without it, “analysis of the subject would be incomplete and inaccurate in the future.”²⁴ Over the next two weeks she spoke twice with the JITF chief, who told her he had given the materials to the J-2’s senior analyst but had heard nothing back.

(U) The analyst then called the J-2’s senior analyst and again recommended that the IC reporting information be published to the entire IC. The J-2 analyst responded that “putting it out there would be playing into the hands of people like Wolfowitz,” that the information “was old” and “only a tid-bit,” asked how did she “know that the information was true,” made a comment about trying to support “some agenda of~eople in the building,” and bucked the issue of publication back to the JITF chief. ~ The JITF took no further action on the recommendation to publish the information, so far as we know.

(U) Meanwhile, the DIA analyst detailed to the Policy Support Office continued to gather and review CIA material on Iraq and al-Qaida. At some point in April or May

23 (U) The Draft Report (page 8) states that this analyst attempted but failed to persuade the JITF Director and the J-2’s senior analyst to publish as an “Intelligence Finding” a July 25, 2002 memo, entitled “Iraq and al-Qaida: Making the Case.” The July memo was an internal document that she wrote in preparation for the SecDef briefing, as discussed more fully below. Nothing in the record known to us indicates any attempt to obtain IC concurrence with the content of the July 25, 2002 memo, nor was there any requirement to do so. Comments to that effect in the Draft Report seem to be a mistaken reference to the earlier effort, in the spring of 2002, to persuade the IC to publish intelligence reports the analyst had found about Iraqi training provided to Bin Laden. The Draft Report claims that “OUSD(P) proceeded to disseminate” the briefing to the SecDef despite being “unsuccessful in convincing the Intelligence Community to publish the alternative intelligence assessments as an Intelligence Finding.” This claim is wrong. There was no attempt to get the IC to publish “alternative intelligence assessments,” there was no requirement to do so, there was no requirement for IC concurrence on the briefing the DSD had directed to be given to the SecDef, and neither the July memo nor the August 2002 briefing contained any “alternative intelligence assessment.”

24 (U) Memorandum for the Record (17 April 2002), USDP Congressional Correspondence November

02-February 04, Tab 17.

25 (U) Ibid. Judging from this response, the J-2’ s senior analyst may have been unfamiliar with DIA’ s DI Policy No. 005 (5 June 2001), the first sentence of which states, “Curiosity and integrity are the hallmarks of good analysis.”

2002, she became aware of the broader work by the PCTEG on various terrorist organizations.²⁶

3. The Deputy Secretary’s Tasking to Brief the Secretary of Defense (U)

(U) Soon thereafter, in approximately July 2002, the DSD initiated the third strand of work relevant here — the strand that resulted in the activities labeled as “inappropriate” in the Draft Report. Specifically, the DSD directed his Special Assistant²⁷to prepare a briefing for the Secretary of Defense on Iraq and links to al-Qaida, based on a review “in a different framework” of IC reports on connections between al-Qaida and Iraq.²⁸In particular, this review was motivated by the issue of whether there was any a priori reason to believe that ideological opponents, (e.g., secular Iraqi Baathists and Islamic extremists) would never cooperate against a common foe. By this point in time, the DSD~s Special Assistant, the DIA analyst detailed to the Policy Support Office, and the single remaining member of the PCTEG had all become aware of the separate but related work of each. Accordingly, the three of them collaborated in preparing the briefing for the Secretary of Defense as directed by the DSD.

(U) The record does not support the Draft Report’s assertion (page 12) that the Deputy Secretary asked for an “intel briefing” when he tasked his Special Assistant to prepare the briefing for the Secretary on Iraq and al-Qaida. That characterization only appears in an internal e-mail, the author of which was not present when the Deputy gave the tasking and had no personal knowledge of how the Deputy in fact formulated his instructions.²⁹

(U) The Report makes much of an internal July 25, 2002 memo entitled “Iraq and al-Qaida: Making the Case.” This memo is dated after its author, the DIA detailee to Policy Support, learned of DSD’s instruction to his Special Assistant to prepare the briefing for the Secretary of Defense³° and, according to its author, was done preliminary

26 (U) Memorandum for the Record (30 October 2002), USDP Congressional Correspondence November

02-February 04, Tab 18.

27 (U) DSD’s Special Assistant at the time was an individual detailed to DSD from the Policy organization. At all relevant times this Special Assistant reported directly to, and took direction exclusively from, the DSD. At no time did the Special Assistant work in or take direction from the OSP, the PCTEG, or the Policy Support Office.

28 (U) Explanatory Note to E-Mail of 7/22/02, USDP Congressional Correspondence November 02-

February 04, Tab 17.

29 (U) Ibid.

30 (U) E-mail dated July 22, 2202, USDP Congressional Correspondence November 02-February 04, Tab

17.

to that briefing.³’ The Report asserts (page 6) that this memo constituted an “OUSD(P)... alternative intelligence assessment.”³²The Report claims that there was then a “translation of that alternative intelligence assessment” into the briefing for the Secretary of Defense, which “translation” the Draft Report characterizes (page 6) as an “Intelligence Activity, and more specifically, Intelligence Production” on the part of

OUSD(P).

(U) To the contrary, the July 25, 2002 memo was not an “OUSD(P)” assessment of any sort, let alone an “alternative intelligence assessment.” Nor was it an “Intelligence Finding” as the Draft Report misleading implies (page 6). It was, rather, a staff-level memo containing only two introductory paragraphs of commentary, followed by a list summarizing various IC reports on contacts between Iraq and al-Qaida.

(U) The Draft Report erroneously asserts (page 9) that the memo described these as “known” contacts. It does not. The phrase “known contacts” does not appear in the memo.

(U) The two introductory paragraphs of the July 25 memo read as follows:

(U)³³Some analysts have argued that Usama Bin Laden would not cooperate with secular Arab regimes such as Iraq because of differences in ideological and religious beliefs. Reporting indicates otherwise. In fact, a body of intelligence reporting for over a decade from varied sources reflects a pattern of Iraqi support for al-Qaida activities. The covert nature of the relationship makes it difficult to know the extent of that support. Moreover, intelligence gaps exist because of ... Iraq’s need to cloak its activities, thus preventing collection of information on additional contacts between Iraq and aJ-Q aid a.

(U)³⁴Published intelligence analyses continue to suggest that ties between Iraq and al-Qaida are not “solid” or “provable.” Intelligence

³¹(U) Letter from USDP to Hon. Pat Roberts (June 29, 2004), USDP Congressional Correspondence March 04-August 04, Tab 30.

32 (U) In contrast, the DIA Senior Intelligence Analyst in the JITF-CT said that the memo had “no intelligence value” because, in the words of the Draft Report, it “contradicted the Intelligence Community assessments. . . .“ (Draft Report page 9).

~ (U) The original version of this paragraph was classified. The classified information has been omitted and the paragraph declassified accordingly.

~ (U) The original version of this paragraph was classified because of content in the bullets that followed it. Those bullets have been omitted here, and the paragraph declassified accordingly.

assessments do not require juridical evidence to support them. Legal standards for prosecution needed in law enforcement do not obtain in intelligence assessments, which look at trends, patterns, capabilities, and intentions. Based on these criteria, the following information clearly makes the case for an Intelligence Finding — that Iraq has been complicit in supporting al-Qaida terrorist activities.³⁵

(U) The Draft Report does not define the term “intelligence assessment,” and we are not aware of a commonly accepted definition. But it is apparent that the above-quoted paragraphs are merely making an argument that the Intelligence Community should make an “Intelligence Finding” that Iraq had been complicit in supporting alQaida terrorist activities. Considering the far more explicit statements to Congress about Iraqi assistance to al-Qaida by the Director of Central Intelligence (DCI) himself, discussed below, the quoted comments by DIA’s detailee to Policy hardly seem extreme. In any case they do not rise to the level of an “intelligence assessment” by the “OUSD(P)” or an “Intelligence Finding” by anyone.

(U) The Draft Report asserts (page 8) that “OUSD(P) disseminated alternative intelligence assessments without Intelligence Community consensus to senior decision-makers.” The Draft Report asserts (page 8) that OUSD(P) should have followed procedures contained in DIA’s DI Policy No. 005 (5 June 2001), which allegedly “detailed appropriate methods within Defense Intelligence for addressing alternative judgments in those rare instances where consensus could not be reached.”

(U) These assertions are wrong. Apart from the fact that the work was not “OUSD(P)” assessments and not in any case “intelligence assessments,” the Draft Report ignores the fact that the Deputy Secretary had asked for a critical reading by non-IC staff members of assessments already provided by the IC. He had not asked for an alternative intelligence judgment and specifically directed that a “consensus” with the IC was not the purpose of this work. As the Deputy wrote in a memo after the briefing to the Secretary:

“That was an excellent briefing. The Secretary was very impressed. He asked us to think about some possible next steps to see if we can illuminate the differences between us and the CIA. The goal is not to produce a consensus ₆product, but rather to scrub one another’s arguments” (emphasis in original).³

(U) It would have been contrary to the Deputy’s direction, not to say futile, for the staffers doing this work to have sought an IC consensus on what was specifically

~ (U) The full text of the July 25, 2002 memo is attached as Tab 2 to Letter from USDP to Hon. Pat

Roberts (June 29, 2004), USDP Congressional Correspondence March 04-August 04, Tab 30.

36 (U) Memo from Paul Wolfowitz to Tina Shelton, et al. (8 August 2002), USDP Congressional

Correspondence November 02-February 04, Tab 17.

intended as a critique of IC work, not as a competing “intelligence assessment.” Yet the OIG apparently believes that it would have been more appropriate for these staff members to have disregarded the Deputy’s direction.

(U) Even if the objective had been to obtain an “alternative intelligence judgment” from the IC, which the Draft Report inexplicably seems to say was or should have been the case, neither DI Policy No. 005 nor DI Policy No. 004 (also cited by the Draft Report) provides any procedure whatever for the DIA’s customers to obtain such an alternative judgment. Both documents are confined solely to situations in which a DIA analyst, working as such within DIA, wishes to put forward an alternative analysis or alternative judgment through DIA’s chain of command. In the present case, one of the individuals responding to the Deputy’s tasking had no connection with DIA at all, and the other two were working in policy positions on detail to OUSD(P). There is no factual or legal basis for the Draft Report’s assertion that these internal DIA policies continued to apply to these detailees while assigned to OUSD(P). The full texts of these internal DIA policies are attached at Appendix A.

(U) The Report claims (page 8) that the DIA detailee who wrote the July 25, 2002 memo “requested first from the Director of the Joint Intelligence Task Force for Combating Terrorism (JITF-CT) and then the Joint Staff J2’s Senior Analyst to publish the alternative intelligence assessment as an ‘Intelligence Finding,” rather than using “the standard process of coordinating to obtain consensus from the Intelligence Community” or to “follow the procedures for developing an Alternative Judgment.” Apart from the mischaracterization of this memo as an “alternative intelligence assessment,” the Draft Report lends great weight to this supposed failure in obtaining IC concurrence, stating (page 8) that “OUSD(P) proceeded to disseminate the August 2002 briefing” to the Secretary though having been “unsuccessful in convincing the Intelligence Community to publish the alternative intelligence assessments as an Intelligence Finding.”

(U) As noted above, this comment may be a mistaken reference to an earlier unsuccessful attempt by the DIA detailee to persuade the JITF to publish intelligence reports she had found on certain training provided to Bin Laden by Iraqi services. Whether or not the memo’s author attempted to coordinate it with the JITF or J-2, there was no requirement to do so since the memo was an internal Policy staff product done in preparation for a briefing that the DSD had directed his staff to prepare for the Secretary of Defense.

(U) It bears emphasis that the DSD gave this direction to his staff, not to the Intelligence Community, as discussed above. Presumably the OIG has interviewed the former DSD to explore his reasons for so doing, though the Draft Report does not elucidate this. The written record seems clear, however, that the DSD was not seeking to have the IC publish an “Intelligence Finding” and was expressly not trying to produce a

consensus product with the IC. Rather, he wanted a critique from a policy perspective of information already provided by the Intelligence Community, followed by an exchange of views with the IC to see how the various arguments might hold up in the give and take of vigorous debate.

B. Draft Briefing to the Secretary of Defense (U)

(U) The briefing, marked “draft,” was given to the Secretary on August 8, 2002~~ and became the first of three versions of the briefing as explained below, all of which were marked as “Draft” or “Draft Working Papers.”³⁸Entitled “Assessing the Relationship Between Iraq and al Qaida,” the briefing summarized existing intelligence products and traffic on contacts between Iraq and al-Qaida. The briefing asked but did not directly answer the following “Key Questions”:

• (U) “What is the probability that there are contacts between Iraq and al Qaida?”

• (U) “What is the probability that there is cooperation regarding such support functions as finances, expertise, training and logistics?”

• (U) “What is the probability that Iraq and al Qaida actually coordinate on decisions or operations?”

• (U) “What is probability that if a relationship existed, Iraq and al Qaida could conceal its depth and characteristics from the United States?”

The briefing then identified various areas of activity in which Iraq and al-Qaida might have an incentive to cooperate, and for each area summarized the available intelligence relating to Iraq’s and al-Qaida’s actions in those areas over time.

(U) One slide entitled “What Would Each Side Want From a Relationship?” lists several categories of potential Iraqi and al-Qaida objectives that each side might help the other in fulfilling (e.g., training, financing, disruption of Kurdish opposition, etc.). It is specifically in regard to these categories that the briefing slide stated “Intelligence indicates cooperation in all categories; mature, symbiotic relationship.”

~ (U) Memo from Paul Wolfowitz to Tina Shelton, et al. (8 August 2002), USDP Congressional

Correspondence November 02-February 04, Tab 17.

38 (U) All three versions of the briefing are attached to Letter from USDP to Hon. Carl Levin (25 March

2004), USDP and Senator Levin Correspondence, November 03-July 05, Tab 9.

(U) The Draft Report (page 6) misquotes this slide by transforming the subjunctive question in the slide’s title (“what would each side want...?”) into an unconditional assertion of “what each side wants from a relationship.”

(U) Contrary to the Draft Report’s mischaracterizations (e.g., page 8), the briefing did not assert that intelligence indicated cooperation in all categories of possible endeavor or a mature, symbiotic relationship in all respects, and “OUSD(P)” most certainly never so contended. No category listed on this slide, and nothing elsewhere in any version of the draft briefing, referred to cooperation in the conduct of specific terrorist operations or to cooperation in operations of any sort.

(U) Here and throughout, the Draft Report misstates what the draft briefing said. It overstates the briefing’s caveated observations as “assessments” and “conclusions,” always arbitrarily attributed to “OUSD(P).”

(U) The whole thrust of the draft briefing was to examine the question, in response to DSD’ s tasking, whether existing intelligence might suggest alternative interpretations if one assumed that Iraq and al-Qaida might be willing to cooperate in a relationship that both would have compelling reasons to hide, and to ask what each side might want from such a relationship.

(U) The question was pertinent because a contrary assumption underpinned a considerable part of the IC analysis, namely, that Iraq’s secular Baathist regime and Islamic extremists such as al-Qaida would not cooperate because of their ideological and religious differences. The Draft Report fundamentally errs in failing to review the draft briefing in the light of its purpose — to respond to DSD’s request for an alternative view based on an alternative assumption.

(U) Each version of the draft briefing included a slide entitled “Findings.” None of these “findings” asserted cooperation between Iraq and al-Qaida in all possible categories of endeavors or a mature relationship in general. The “findings” in their entirety were as follows:

• (U) “More than a decade of numerous contacts”

• (U) “Multiple areas of cooperation”

• (U) “Shared anti-US goals and common bellicose rhetoric

-- Unique in calling for killing of Americans and praising 9/11”

• (U) “Shared interest and pursuit of WMD”

• (U) “[One indication of]³⁹[One possible indication of]⁴° [Some indications of possible]⁴¹Iraqi coordination with al Qaida specifically related to 9/11”

• (U) “Relationship would be compartmented by both sides, closely guarded secret, indications of excellent operational security by both parties”

(U) The reference to possible “coordination with al Qaida specifically related to 9/li” was at no time presented as a conclusion that Iraq and al-Qaida had in fact cooperated in regard to the 9/11 attacks.

(U) Furthermore, both versions briefed outside the Defense Department were caveated by the word “possible” in reference to “coordination.” And all three versions of the draft briefing included an additional caveat, in a slide preceding the “Findings” slide, stating that ‘~fragmentary reporting points to possible Iraqi involvement” in 9/11 and previous al-Qaida attacks (emphasis added).

(U) These caveated statements in the draft briefing were not “OUSD(P)” “assessments” and were not presented as such at any of the three presentations of the briefing.

(U) The Draft Report errs in its repeated assertion (e.g., page 7) that “OUSD(P) assessed the Iraq — al-Qaida relationship as having a higher degree of cooperation than those conclusions supported by the Intelligence Community.” As discussed above, the draft briefing was more conditional and less certain in its discussion of “possible” cooperation than the Draft Report asserts.

(U) On the other hand, the DCI’s statements on the subject — which the Draft Report does not address — were more robust than the OIG admits. The Draft Report attempts to portray a wide gulf between the draft briefing’s observations and the IC’s assessments by quoting from IC products stating that there are “no conclusive signs of cooperation on specific terrorist operations” and no “compelling evidence demonstrating direct cooperation” (page 7). But, as discussed, the draft briefing never asserted that there was any operational relationship or any cooperation on specific terrorist operations.

(U) In any event the draft briefing was not an “OUSD(P)” assessment of any sort. Nowhere did any version of the draft briefing state that it presented an “OUSD(P)” position or assessment, the USD(P) never approved or represented the draft briefing as an

‘~ (U) Version briefed to the Secretary of Defense.

40 (U) Version briefed to the DCI.

⁴¹(U) Version briefed to the Deputy National Security Advisor.

“OUSD(P)” assessment, the Draft Report cites no facts supporting its repeated assertions to the contrary, and there are none.

C. The Secretary of Defense’s Direction to Brief DCI, Draft Briefing to DCI, CIA Meeting (U)

(U) After receiving the briefing on August 8, 2002, the Secretary of Defense directed that it be given to the DCI, which was done on August 15, 2002 at the CIA.⁴² The USDP attended this meeting and was accompanied by two of the authors of the briefing. At the outset of the meeting the USDP made a statement stressing that this briefing was merely one way of looking at the underlying information, that no one was saying it was necessarily the correct way, and that there were also other ways to view the information. In other words, he made clear that the briefing was for the purpose of discussion and was not presented as an approved OSD or OUSD(P) position.

(U) The draft briefing as given to the DCI did not include a slide entitled “Fundamental Problems with How Intelligence Community is Assessing Information” that was included in the other two versions. This slide criticized the IC for applying an overly strict “juridical” standard in its assessments of the Iraq-al-Qaida relationship, underestimating the importance each side would attach to hiding a relationship, and making an assumption that secularists and Islamists would not cooperate even when they had common interests. It was omitted from the DCI briefing because its critical tone at the DCI-hosted meeting might have distracted from a discussion of the substance.⁴³Even without the omitted slide, however, it was clear from the overall content that the draft briefing was suggesting insufficient attention and analysis by the IC to a number of intelligence reports on contacts between Iraq and al-Qaida — a point that was made explicitly at a subsequent meeting at CIA on August 20, 2002, discussed below.

(U) The reference in the briefing to possible Iraqi coordination with al-Qaida related to 9/11 was based on a report from the Czech intelligence service that future 9/11 highjacker Mohammad Atta had met with the Prague chief of the Iraqi Intelligence Service in April 2001. All three versions of the draft briefing, including the one given to the DCI, had a slide entitled “Summary of Known Iraq-al Qaida Contacts, 1990-2002” that included the statement “2001: Prague uS Chief al-Ani meets with Mohammad Atta in April.”

( ) Whether or not it was an overstatement to describe the reported Atta meeting as a “known contact,” the fact is that at the time of this briefing the Czech intelligence service stood firmly by its report and apparently

42 (U) SSCI Report, p. 362.

⁴³(U) Letter from USDP to Hon. Carl Levin (25 March 2004), USDP and Senator Levin Correspondence, November 03-July 05, Tab 9.

In contrast, the CIA report cited at page 7 of the Draft Report describes the reporting on the al”~’~ ~ as “..

~ 7, states that the Atta

me - _____________i.” But at no time

relevant to this Project did the US Intelligence Community articulate and disseminate any conclusive coordinated judgment that the reported Atta meeting did not occur.

(U) In any case all versions of the draft briefing merely spoke of an “indication” of “coordination” regarding 9/11 in regard to this alleged meeting, both versions presented outside the Defense Department added the further caveat of “possible,” and no version of the draft briefing asserted that Iraq and al-Qaida actually cooperated operationally or otherwise in regard to the 9/11 attacks.

(U) Furthermore, during all times relevant to this Project the question of the reported Atta meeting was well known and vigorously discussed throughout USG policy and intelligence circles with responsibility for Iraq. There can be no doubt that all recipients of the draft briefing, and most particularly the Secretary of Defense, the DCI, the Deputy National Security Advisor and the Vice President’s Chief of Staff, were aware of the controversy surrounding the alleged meeting. They all were recipients of the IC’s judgments on this and related matters, both before and after receiving the draft briefing. There is no factual basis whatever to suggest that any of them would have been misled by anything about this meeting in any version of the draft briefing, or would have misunderstood the draft briefing to be some sort of “intelligence assessment” by

OUSD(P).

(U) The DCI reportedly found the briefing “useful.”⁴⁴The DCI asked the OUSD(P) staffers to speak with the CIA’s NESA and CTC experts on Iraq and terrorism. As a result, the two OUSD(P) staffers who briefed the DCI were invited to attend an August 20, 2002 meeting of analysts from the CTC, NESA, the National Security Agency and the DIA who convened to discuss ongoing intelligence community work assessing Iraq’s links to terrorism. At the meeting the OUSD(P) staffers pointed out various intelligence reports that had not been included in finished intelligence products and suggested that such reports should be included. Some of their suggestions were adopted and some were not.

(U) The Draft Report notes (page 10) that in this meeting the “CIA was even willing to footnote its report with the OUSD(P) conclusions that differed from the report’s findings.” In fact, there was no offer to footnote “OUSD(P) conclusions,” and in any case there were no “OUSD(P) conclusions” on the matter at hand, hence none to

~“ (U) Memo Entitled “Quick Points on the Policy Team’s Visit with DCI” (16 August 2002), USDP and Senator Levin Correspondence, November 03-July 05, Tab 9.

DIA report, also cited at

footnote. Also, the OUSD(P) staffers in attendance did not decline footnotes because they were “unable to speak for Defense Intelligence” as the Draft Report (page 10) puts it, although in fact they were not. The actual exchange was simply this: One of the OUSD(P) staffers (the DIA analyst to the Policy Support Office), when asked to prepare footnotes on the issues with which she disagreed, declined to do so, stating that “I was an employee in Policy, not wearing an intelligence hat. I could only ask why reporting was not included in finished products and ... make recommendations to include it.”⁴⁵

(U) In its unanimous report on pre-war intelligence issues in July 2004, the Senate Select Committee on Intelligence stated that all attendees of the August 20, 2002 meeting “interviewed by the Committee staff (eight of the twelve individuals) agreed that the OUSDP staffers were not given special treatment[,] . . . their attendance contributed to a frank exchange of opinions” and they “played by IC rules. .,“~ The Committee Report also noted more generally that:

“In some cases, those interviewed stated that the questions had forced them to go back and review the intelligence reporting, and that during this exercise they came across information they had overlooked in initial readings. The Committee found that this process — the policymakers probing questions — actually improved the Central Intelligence Agency’s ... products.”⁴⁷

D. Deputy National Security Advisor’s Request, DSD’s Direction, Draft Briefing to Deputy National Security Advisor (U)

(U) The Draft Report mischaracterizes these events as “Dissemination of OUSD(P)’s Alternative Intelligence Assessment to the White House” page 10). What transpired is this:

(U) Following a reference to the briefing at a Deputies Committee meeting in August 2002, the Deputy National Security Advisor requested to receive the briefing. The Deputy DCI was a designated member of the Deputies Committee, and he or his designee consistently attended its meetings. On the morning of September 16, 2002, the Deputy Secretary’s office instructed the OUSD(P) staffers who had helped prepare the draft brief to present it to the Deputy National Security Advisor and the Vice President’s Chief of Staff. They did so the same day at a meeting hosted by the Deputy National Security Advisor in the Situation Room, with the Vice President’s Chief of Staff attending for at least part of the meeting.

~ (U) Memorandum for the Record (30 October 2002), USDP Congressional Correspondence Nove~n her 02-February 04, Tab 17.

46 (U) SSCI Report, pp. 362, 363.

~ (U) SSCI Report, p. 34.

(U) The Draft Report fails to mention that the OUSD(P) staffers gave the September 16 briefing because they were instructed to do so by the Deputy Secretary’s office in response to the Deputy National Security Advisor’s request. The Draft Report does correctly state (page 29) that there was no requirement for the DCI to be informed of this meeting. One might reasonably observe that there was no requirement because the meeting was not an intelligence meeting.

(U) In any case, this version of the draft briefing, just as the previous two versions, contained no intelligence assessment and was not presented as an official OUSD(P) position. It was presented not as an intelligence briefing but as an alternative assessment of IC reports, just as the prior two versions of the briefing.

(U) The Draft Report states (page 11) that this version of the draft briefing included a “previously unseen” slide entitled “Facilitation: Atta Meeting in Prague.” The Draft Report fails to point out that the slide was “previously unseen” because it did not previously exist. The Draft Report incorrectly asserts that this new slide presented the alleged Atta meeting “as fact” (page 11). Nowhere does the slide describe the meeting as “fact.” To the contrary, the slide repeatedly uses phrases such as “Czech service reports that Atta visited . . . ,“ “despite press reports of conflicting information, Czech Interior Minister ... stands by previous Czech ... reporting,” “Atta reportedly held meetings...,” and “Atta reportedly arrives in Prague... ,,48

(U) Furthermore, the attendees at this version of the draft briefing were well informed senior officials who had access to all the IC’s most highly classified and compartmented information on the subject of the alleged Atta meeting. The Deputy National Security Advisor and the Vice President’s Chief of Staff certainly were familiar with the debate in the US Intelligence Community on this subject. It is ludicrous to suggest that they would have mistaken this slide or anything else in the draft brief as firm assertions of fact, much less as “intelligence assessments” by “OUSD(P)” or anyone else.

E. DCI’s Congressional Statements on Iraq and al-Qaida (U)

(U) The Draft Report partially quotes from several IC reports, casting doubt on the existence of any significant cooperation between Iraq and al-Qaida, in asserting that the work under review overstated the degree of cooperation and hence “OUSD(P)” did not provide “the most accurate analysis of intelligence” (page 11). As noted above, the responsibility to provide “the most accurate analysis of intelligence” rests with the IC, not OUSD(P). More importantly, senior decision-makers already had the IC’s reports and assessments on Iraq and al-Qaida and thus already had “the most accurate analysis of

48 (U) USDP and Senator Levin Correspondence, November 03-July 05, Tab 9.

intelligence” -- if one accepts, as the Draft Report seems to do, that the IC’s assessments are the “most accurate”.

(U) It is puzzling, therefore, that the Draft Report fails to discuss some of the most authoritative articulations of the IC’s analysis on Iraq and al-Qaida — the vetted, coordinated correspondence and testimony by the DCI himself to the Congress. On October 7, 2002, the DCI wrote to SSCI Chairman Graham, responding to various questions raised in connection with the forthcoming debate on a joint resolution to authorize military action against Iraq. Regarding questions about Iraqi links to al-Qaida, the DCI wrote that Senators could draw from the following points for unclassified discussions:

• Our understanding of the relationship between Iraq and al-Qa’ida is evolving and is based on sources of varying reliability. Some of the information we have received comes from detainees, including some of high rank.

• We have solid reporting of senior level contacts between Iraq and atQa’ida going back a decade.

(U) By comparison, the draft briefing referred to “more than a decade of numerous contacts. The DCI’s letter continued:

• Credible information indicates that Iraq and al-Qa’ida have discussed safe haven and reciprocal non-aggression.

(U) The draft briefing referred to “safe haven of last resort” as an objective that alQaida would want from a relationship with Iraq. The DCI’s letter continued:

• Since Operation Enduring Freedom, we have solid evidence of the presence in Iraq of al-Qa’ida members, including some that have been in Baghdad.

(U) The draft briefing said that “Iraq Has Provided Safe Haven for Key Terrorists,” among them al-Qaida members, including some in Baghdad. The DCI’s letter continued:

• We have credible reporting that al-Qa’ida leaders sought contacts in Iraq who could help them acquire WMD capabilities. The reporting also stated that Iraq has provided training to al-Qa’ida members in the areas of poisons and gases and making conventional bombs.

(U) The draft briefing said that Iraq and al-Qaida had a “shared interest and pursuit of WMD,” that “CBRN” would be an al-Qaida objective, and that al-Qaida had sought bomb-making assistance. The DCI’s letter continued:

• Iraq’s increasing support to extremist Palestinians, coupled with growing indications of a relationship with al-Qa’ida, suggest that Baghdad’s links to terrorists wilt increase, even absent military action.⁴⁹

(U) In a prepared statement to the SSCI on February 11, 2003, DCI Tenet said:

Iraq has in the past provided training in document forgery and bomb-making to al-Qa’ida. It also provided training in poisons and gasses to two al-Qa’ida associates; one of these associates characterized the relationship he forged with Iraqi officials as successful. Mr. Chairman, this information is based on a solid foundation of intelligence. It comes to us from credible and reliable sources.⁵°

(U) At a Senate Armed Services Committee hearing on February 12, 2003, the DCI stated:

IWIe also know from very reliable information that there’s been some transfer of training in chemical and biologicals Esici from the Iraqis to at Qaeda.⁵’

(U) From these statements by the DCI on behalf of the Intelligence Community, it is clear that the IC “consensus” at the time ascribed considerably more “maturity” and “symbiosis” to the relationship between Iraq and al-Qaida than depicted in the Draft Report. It is also clear that the Draft Report significantly overstates the degree and significance of inconsistencies between the IC consensus and the draft briefing’s observations. In any case the draft briefing was nothing more than a draft, it was not an “intelligence assessment,” and it was not an “OUSD(P)” assessment or conclusion.

~ (U) Letter George Tenet, DCI, to Hon. Bob Graham, Chairman SSCI (7 October 2002), in E-Mailfrom Michael H Mobbs (OUSDP) to Charles E. Edge (OIG) (7 February 2006), at Tab C.

~° (U) “Administration Statements on Iraq Training al Qa’ida in Chemical and Biological Weapons,” attached to Press Release by Senator Carl Levin Re: Levin Says Newly Declassified Information Indicates Bush Administration’s Use of Pre-War Intelligence Was Misleading (6 November 2005), in EMail from Michael H Mobbs (OUSDP) to Charles E. Edge (OIG) (7 February 2006), at Tab C.

⁵¹(U) Ibid.

V. DISCUSSION (U)

A. Why are Lawful and Authorized Activities Nevertheless Called

“Inappropriate”? (U)

(U) The Draft Report concludes that the activities reviewed in this Project were lawful and authorized (pages ii, 4, 13). It states that within the authority conferred by Title X, Section 113 of the United States Code, “the Secretary owns the DoD Directives governing (among others) Intelligence and Policy, and as long as Executive Orders or other legal statutes are not violated, he has the latitude to interchange roles and responsibilities” (page 34).

(U) Despite these conclusions, the Draft Report asserts that these same activities were “inappropriate,” in the OIG’s opinion, because the “OUSD(P)” “products did not clearly show the variance with the consensus of the Intelligence Community and were, in some cases, shown as intelligence products” (page 4).

(U) It is somewhat difficult to understand how activities that admittedly were lawful and authorized (in this case by either the Secretary or the Deputy Secretary of Defense) could nevertheless be characterized as “inappropriate” -- particularly considering OIG’s concession that the Secretary (and by logical extension the Deputy) may interchange roles and responsibilities within DoD provided no statutes or executive orders are violated. The Draft Report points to no laws, executive orders, DoD directives, DoD instructions or DoD publications that provide any guidelines for what is “appropriate” in this case, except for the Secretary’s broad mandate under Title X. That mandate leads to a conclusion that the activities reviewed were “appropriate.”

(U) The Draft Report is spare of analysis on why its reaches the opposite conclusion. The argument seems to be as follows:

• (U) DIA detailees to OUSD(P) reviewed the same intelligence information that the IC had used when drawing IC judgments about links between Iraq and al-Qaida. This was appropriate for policy formulation (page 12).

• (U) Appropriate policy formulation, however, “evolved into Intelligence Analysis and eventually culminated in the Intelligence Activity of Intelligence Production with the creation of alternate intelligence assessments and dissemination when the briefing was provided to the Secretary of Defense, DCI, and members of the Office of the Vice President and National Security Counsel” (page 12).

• (U) This supposed “evolution” was inappropriate because it led to performance by “OUSD(P)” of “intelligence functions that are the responsibility of Defense Intelligence” (page 14), the work products “did not clearly show the variance with

the consensus of the Intelligence Community” (page 4), and the work products “were, in some cases, shown as intelligence products” (page 4).

• (U) If “OUSD(P)” did not consider the IC’s existing “judgment” about Iraq and alQ aida to be correct, “OUSD(P)” should have used “existing procedures” to get a second IC “judgment” by requesting “from the Defense Intelligence community an Alternative Judgment” on that subject (pageS 13-14) — instead of participating in an OSD critique of the existing IC judgment as directed by the DSD. Such “existing procedures” are said to be found in two internal DIA policies cited in the Draft Report (DI Policy Nos. 004 and 005).

(U) It is apparent from the above summary that the Draft Report’s conclusions about “inappropriate” activities rest heavily on internal DIA policies dealing with alternative IC assessments and judgments, as well as Intelligence Community concepts such as “Intelligence Activities,” “Intelligence Production,” “Intelligence Analysis,” and “intelligence assessments.” An examination of the DIA policies and relevant IC concepts shows that they do not apply to the activities reviewed here. Thus the assertion that the activities were “inappropriate” cannot withstand analysis.

(U) Before turning to the analytical errors in the Draft Report, however, we respectfully point out that the specific reasons on which the Draft Report rests its finding of “inappropriateness” do not bear scrutiny.

(U) First, the Draft Report claims that the work products were inappropriate because they “did not clearly show the variance with the consensus of the Intelligence Community.” This fundamentally mischaracterizes the purpose and nature of the work. The central purpose of these activities was to look critically at existing IC work and offer a different way of understanding the IC information. Each version of the draft briefing made this clear. The senior decision-makers briefed on this work (one of whom was the DCI himself) did not need to be told that it was at variance with the IC in some respects; that was inescapably obvious. There are no facts to suggest that any of them drew any conclusions or made any decisions whatsoever solely on the basis of the draft briefing, without taking IC views into consideration.⁵²

(U) Furthermore, there was no requirement to specify in a draft work product, not offered as a proposed action item, how it might vary from IC views. The situation would

52 (U) It was not the place of OUSD(P) in any event to articulate what the IC “consensus” was, which would have been the first step in “clearly show[ing] the variance” as the Draft Report asserts should have been done. It was up to the IC to articulate its consensus, if it had one. The Draft Report itself shows the pitfalls of trying to articulate an “IC consensus” for the IC. The Draft Report purports to describe such a consensus but utterly fails to mention the DCI’s vetted, cleared statements to Congress on the Iraq-alQaida relationship. Those statements do not support the Draft Report’s characterization of the IC “consensus.”

have been different if the draft briefing were put forward in support of some proposed action or decision, for example, a proposal that the President make a speech to the Nation describing a relationship between Iraq and al-Qaida. In such a case, the matter would have been discussed, at the least, by the Deputies Committee. All interested agencies would have been asked to provide their views, in particular their comments on the draft briefing and any other material offered in support of or against the proposed speech. The IC would have had ample opportunity to articulate how its views did or did not vary from the draft briefing. There would have been no need for “OUSD(P)” to do that; indeed, the IC would no doubt have objected strenuously to the idea of having another agency describe how its views might vary from those depicted in the draft briefing. Obviously, nothing of the sort happened here.

(U) Second, the Draft Report asserts that the work was “inappropriate” because some of it was “shown as intelligence products.” There are no facts whatsoever to support this statement. The Draft Report only gives one example, the July 25, 2002 internal staff memo (done in preparation for the draft briefing), discussed at length in Part IV above. That memo argued that the IC had sufficient information to make an intelligence finding that Iraq had been “complicit in supporting al-Qaida terrorist activities.” The Draft Report mischaracterizes this memo as an “OUSD(P)” intelligence assessment. In fact it was nothing more than a staff member’s opinion that the IC should make an intelligence finding.

(U) Third, the Draft Report considers the work reviewed inappropriate because it amounted to “intelligence functions that are the responsibility of Defense Intelligence.” We explain below why the work was not “intelligence functions.” But even accepting that characterization for discussion purposes only, the Draft Report in this respect contradicts its own admission that the Secretary “has the latitude to interchange roles and responsibilities” in managing the Department so long as no statutes or executive orders are violated. The Draft Report fails to explain why it was inappropriate for the Secretary and Deputy Secretary to exercise that latitude in this case. If the OIG believes the Deputy inappropriately used his latitude to assign this work to non-IC staff members, and the Secretary and Deputy misused their latitude to direct that those staff members share this work outside the Department, it is incumbent on the OIG to say so directly and to explain why it holds this opinion. It is not sufficient for the OIG simply to fault “OUSD(P)” with engaging in “inappropriate” behavior because two Policy staffers did as told by the Secretary and Deputy, and let it go at that.

B. DIA’s DI Policy Nos. 004 and 005 Do Not Apply to Non-IC Offices Directed by Senior DoD Leaders to Critique Intelligence Community Work (U)

(U) The Draft Report cites Policy Nos. 004 and 005 developed by DIA’s Directorate for Analysis and Production. These internal policies set out guidelines and procedures for DIA analysts who wish to propose, respectively, an alternative analysis or

an alternative judgment when they believe that they cannot reach a consensus with other intelligence analysts on a particular issue. The Draft Report erroneously characterizes these internal DI policies as “the standard process of coordinating to obtain consensus from the Intelligence Community” that the DIA detailees to OUSD(P) should have used in this case (page 8). The Draft Report also erroneously describes these internal policies as the “existing procedures” (page 14) that OUSD(P) should use to “request that an Alternative Judgment be produced by Defense Intelligence” if OUSD(P) believes that the IC is incorrect on a given matter (page 13).

1. The Internal DIA Policies Do Not Apply to DIA Members While Detailed to Policy Positions Outside DIA’s Chain of Command (U)

(U) The texts of these internal DI policies are reproduced in full at Appendix A to these comments. There is nothing in either of them to support the idea that they continue to apply to DIA analysts who are detailed to policy positions and who are tasked to do independent assessments for the express purpose of providing a non-IC critique, or review, of IC views. It is obvious from the texts that they only apply to analysts working within the DIA chain of command and proposing alternative assessments or judgments, in an intelligence capacity, within that chain of command. DI Policy No. 005, for example, provides that “the analyst forwards ... through the immediate Supervisor/Office Senior Intelligence Officer (SIO) to the Group SIO/Research Director (RD). The Supervisors/Office SIOs review ... for format and completeness. The Group S10/RD reviews ... to ensure it accurately describes the competing analyses,” etc. This process has no relevance to a situation such as the present, where the Deputy Secretary specifically directed that he wanted an alternative look at the IC’s work from outside the IC and was not seeking to develop a consensus.

2. The Internal DIA Policies Contain No Procedure for an IC Customer to

Obtain an Alternative IC Judgment, Which in any Case is not What the DSD

Sought Here (U)

(U) Neither of these internal DI policies contains any procedure for an IC customer, such as OUSD(P), to request an “alternative judgment” from the DIA if the customer considers an existing IC judgment to be incorrect. While the Draft Report inexplicably allows that OUSD(P) “is not ... required to await final adjudication or production of an Alternative Judgment from DIA” (page 13), thus raising the question of why the “Alternative Judgment” should be sought at all, the fact remains that these internal DI policies do not provide for a customer to make such a request. One will search the texts in vain for even the slightest hint of such a procedure.

(U) The very notion that a customer should ask the IC for an alternative intelligence judgment if it dislikes the judgment already given is bizarre on its face. Such a request would inevitably bring down a firestorm of criticism that the customer was attempting to “politicize” intelligence or “pressure” the intelligence analysts into

changing their assessments. In any event, the Deputy Secretary in the present matter expressed no wish for an “alternative judgment” from the IC, which is undoubtedly why the staffers responding to his tasking did not seek one. And he expressly directed that the objective of the work was not to develop a consensus product but rather to see how competing arguments might stand up in an exchange of views with the IC.

3. The Internal DIA Policies Were Not Coordinated or Published as Would Have Been Required if Intended to Apply Outside DIA (U)

(U) There is no basis for asserting that the DI internal policies are applicable to DoD as a whole or to OUSD(P) in particular. To the contrary, these policies have not been published; they have not been disseminated to OUSD(P) or, so far as we know, elsewhere in the Department outside DIA; and they have not been presented to OUSD(P) for review or coordination.

(U) Guidance that is intended to have Departmental applicability falls within the requirements of DoD Directive No. 5025.1, “DoD Directives System,” July 27, 2000, as reissued July 14, 2004. Section 4.1 of this directive articulates a DoD policy to maintain “a single, streamlined, uniform system governing the preparation, coordination, approval, publication, dissemination, implementation, and internal review of DoD issuances....” Proposed DoD issuances “shall be formally coordinated to solicit the views of the Heads of the DoD Components” (Section 4.4). All DoD issuances “must be coordinated with the General Counsel, DoD, the Inspector General, DoD, and the Director of Administration and Management” (Section 4.4.1). The Heads of DoD Components “shall review and coordinate on proposed DoD issuances relevant to their missions” (Section 5.4).

(U) Nothing of the sort was done with regard to DI Policy Nos. 004 and 005. They have no applicability to OUSD(P). They are not “existing procedures” that OUSD(P) should have, or could have, followed in the present matter. The Draft Report’s recommendation that they be followed as “existing procedures” in the future is unfounded and inappropriate.

C. “Intelligence Activities” Constitute a Process Using All Key Elements of Intelligence Work By Intelligence Agencies (U)

(U) As the guidance cited by the Draft Report (page 4-5, Appendix H) and other relevant authorities make clear, “Intelligence Activities” involve the entire process by which intelligence agencies turn information into a product that intelligence consumers can use. They do not encompass the type of work reviewed here.

(U) In asserting otherwise, the Report relies primarily on DoD Directive No. 5240.1, “DoD Intelligence Activities, April 25, 1988, and DoD Directive No. 5105.21, “Defense Intelligence Agency (DIA),” February 18, 1997. Of these, only DoD Directive

No. 5240.1 (Section 3.1) contains a definition of “Intelligence Activities” which is as follows:

“Intelligence activities. The collection, production, and dissemination of foreign intelligence and counterintelligence by DoD intelligence components authorized under reference (b).”

(U) “Reference (b)” is Executive Order 12333, “United States Intelligence Activities,” December 4, 1981, Section 3.4(e) of which defines “intelligence activities” as “all activities that agencies within the Intelligence Community are authorized to conduct pursuant to this Order.” Section 3.4(f) defines “Intelligence Community and agencies within the Intelligence Community” as “the following agencies or organizations,” among which the Office of the Secretary of Defense and the Office of the Under Secretary of Defense do not appear.

(U) DoD Directive No. 5240.1, Section 3.4, similarly defines “DoD intelligence components” as “[a]ll DoD Components conducting intelligence activities, including” a list of named DoD elements among which, again, the Office of the Secretary of Defense and the Office of the Under Secretary of Defense do not appear. In contrast Section 2.1 of DoD Directive No. 5240.1 does define “DoD Components” to include the Office of the Secretary of Defense. Thus the Directive carefully distinguishes “all DoD Components” from “DoD Components conducting intelligence activities.” In consequence, the Directive’s Section 3.1 definition of “Intelligence Activities” by its terms only encompasses “DoD intelligence components,” not “all DoD Components.”

(U) The above definitions make clear that “Intelligence Activities” constitute a process that entails collection, production “and” (not “or”) dissemination of foreign intelligence or counterintelligence as conducted by intelligence agencies, and not assessments or critiques by non-intelligence offices.

(U) Various definitions in Joint Publication 1-02, Department of Defense

Dictionary of Military and Associated Terms (12 April 2001, as amended through 16

October 2006) (“JP 1-02”) also demonstrate that the term “Intelligence Activities” should

be understood as a process of actions and operations conducted by the Intelligence

Community to produce an intelligence product for consumers. For example, according to

JP 1-02:

“intelligence” means “[t]he product resulting from the collection, processing, integration, analysis, evaluation, and interpretation of available information concerning foreign countries or areas” (JP1-02 at 268);

“intelligence process” means “[t}he process by which information is converted into intelligence and made available to users. The process consists of six

interrelated intelligence operations: planning and direction, collection, processing and exploitation, analysis and production, dissemination and integration, and evaluation and feedback” (JP 1-02 at 270); and

“intelligence community” means “[aJll departments or agencies of a government that are concerned with intelligence activity, either in an oversight, managerial, support, or participatory role” (JP 1-02 at 269).

(U) None of the above definitions accurately describe the critical assessment of IC information by OSD staff members that is the subject of this review.

D. Alternative or Critical Assessments of IC Information and IC Judgments by Non-IC Offices Are Not “Intelligence Activities” (U)

(U) As the above definitions of “Intelligence Activities” and related terms make clear, such activities consist of the entire process of actions and operations conducted by intelligence agencies to produce an intelligence product for consumers. It is incorrect to select one or a few activities that are part of the “intelligence process” and characterize those selected activities as “Intelligence Activities” even when conducted by non-IC policy elements of government.

(U) The definitions of “Intelligence Activities” and related terms do not encompass an alternative or critical analysis, evaluation, interpretation or assessment by a non-IC office, such as OSD or OUSD(P), of information provided by the Intelligence Community. In this context, the “analysis,” etc. is merely an independent review by a non-IC organization, or in the present case by several non-IC OSD staffers, of IC information provided by the IC. In conducting this review, the non-IC organization may even exercise independent judgment about the meaning or significance of the intelligence information provided by the IC. This act of independent judgment by the non-IC organization does not constitute “Intelligence Activities” under any of the above definitions or any common-sense understanding.

(U) The mere fact that the “intelligence process” conducted by the Intelligence Community includes but is not limited to “analysis” and “dissemination” does not mean that a policy organization is conducting “Intelligence Activities” if it independently “analyses” intelligence information provided by the IC and then “disseminates” the results of its analysis. To assert such a proposition is akin to asserting that “cows have four legs and give milk, therefore, all four-legged animals that give milk are cows.”

(U) The Draft Report cites the definition of “Intelligence Production” found in DoD Directive No. 5105.21 in an effort to characterize OUSD(P) activities as “Intelligence Activities.” But the actual definition does not support this argument.

(U) The term “Intelligence Production” as defined in Directive No. 5105.21 does not apply to any activities under review here. Paragraph E2.l.3 of the Directive provides:

“Intelligence Production. The validation, correlation, analysis, and interpretation of information on foreign intelligence and counterintelligence topics, including the use of automated data bases and the presentation and dissemination of the results.”

This definition, just as the related definitions discussed above, makes clear that “Intelligence Production” is the full process of validation, correlation, analysis, interpretation, presentation and dissemination. It is a distortion of the definition to assert that a single activity, such as analysis or interpretation, constitutes “Intelligence Production.”

(U) In the present matter, the draft briefing and work done to prepare it were nothing more than a critical review of intelligence information already produced by the IC. The work presented a fresh assessment of how that information might be understood if certain a priori assumptions about lack of cooperation between secularists and fundamentalists were avoided. At the very least the work under review involved no validation or correlation, as those tasks had already been done by the IC as part of its “Intelligence Production.” The attempt to stretch the definition of “Intelligence Production” to include the critique of IC reports and products by a non-IC office simply does not work.

E. OUSD(P) Did Not Produce or Disseminate “Intelligence Assessments” or “Intelligence Analysis” (U)

(U) The Draft Report asserts (e.g., page 4) that the draft briefing on the relationship between Iraq and al-Qaida and the July 25, 2002 memo preliminary to the briefing were “OUSD(P)” “alternative intelligence assessments,” and that this work “evolved into Intelligence Analysis” (page 12). The work reviewed was not “intelligence assessments” or “Intelligence Analysis” under any reasonable understanding of those terms.

(U) Neither the Draft Report, nor any of the authorities mentioned there or here, defines the term “intelligence assessment.” Nor do they define the term “Intelligence Analysis” despite the Draft Report’s use of capital letters. But extrapolating from the intelligence-related definitions discussed above, it seems reasonable to suggest that “intelligence assessments” and “Intelligence Analysis” are assessments and analysis by intelligence agencies about the meaning and significance of information acquired by them during the six-part “intelligence process” of “planning and direction, collection, processing and exploitation, analysis and production, dissemination and integration, and evaluation and feedback” (JP 1-02 at 270). It follows that “intelligence assessments” and “Intelligence Analysis” are disseminated by intelligence agencies and are clearly

identified as the “assessment” or “analysis” of the issuing agency or intelligence community. Thus, intelligence consumers will know that they have the “assessment” or “analysis” of that agency or community on the matter at hand as opposed to someone else’s assessment or analysis.

(U) Nothing of this sort took place in preparing and presenting the draft briefing in question. As Part IV (Facts) above explains in detail, the July 25, 2002 memo was an internal document done in preparation for a briefing that the Deputy Secretary had directed his Special Assistant and two DIA detailees working in the Policy organization to put together for the Secretary of Defense. The memo did not present any “intelligence assessment” or “intelligence finding” or anything that could reasonably be characterized in that way. The memo did argue that there was a case to support an “Intelligence Finding” that Iraq had been complicit in supporting al-Qaida terrorist activities. But this obviously was a suggestion that the Intelligence Community should make such an “Intelligence Finding,” since neither the memo’s author nor OUSD(P), the Deputy Secretary or the Secretary were capable of making an “Intelligence Finding.”

(U) As Part IV above also explains, the draft briefing likewise contained no “intelligence assessments,” “Intelligence Analysis” or anything that could reasonably be so described. Each version of the draft briefing was marked as “draft” or “draft working papers.” Each time the briefing was given, it was well known to all in attendance that the briefers were not speaking for the Intelligence Community but, to the contrary, were presenting an alternative or critical analysis of information provided by the Intelligence Community. The analysis intentionally took a different approach from some of the IC analysis, because of the Deputy Secretary’s direction to avoid the a priori assumption that secular Baathists and Islamic fundamentalists would never cooperate and to examine how the intelligence information might be understood in the absence of that assumption. It would be preposterous to suggest that the draft briefing was an effort to usurp the role of the IC, or that anyone was misled into believing that the draft briefing purported to express “intelligence assessments” or “Intelligence Analysis” on behalf of the IC or anyone else.

(U) Moreover, whatever the July 25, 2002 memo and the draft briefing may have been, they most certainly were not “OUSD(P)” assessments or conclusions, as the Draft Report repeatedly asserts. As Part IV (Facts) discusses in detail, these work products were never described or presented as an approved OUSD(P) or OSD position, all versions of the briefing were marked “draft” or “draft working papers,” the USDP introduced the draft briefing to the DCI stating that it was merely one way of looking at the underlying intelligence and not necessarily the correct way, and the draft briefing itself was done at the Deputy Secretary’s direction. The draft briefing and work leading to it were not initiated by “OUSD(P),” notwithstanding that two of the three authors happened at the time to be working in the Policy organization on detail from DIA.

(U) The Draft Report seems to argue that the two DIA detailees continued to function as intelligence analysts even though detailed to OUSD(P) and therefore their activities in OUSD(P) “constituted intelligence analysis and in at least several cases, intelligence production, which was not one of USD(P)’s specified functions in DoD Directive 5111.1” (page 6). This contention cannot withstand scrutiny. If it were correct, OSLTD(P) could never obtain intelligence analysts on detail from DIA without committing “inappropriate” “Intelligence Activities.” How to characterize work done by detailees depends on the substance of what they actually do while detailees, not on the nature of their duties in their home agencies. As demonstrated above, the work in question here did not fall within any of the definitions of “Intelligence Activities” and did not constitute “intelligence analysis.”

(U) The Draft Report also seeks to support its claim that OUSD(P) produced “alternative intelligence assessments” by referring to “confirmation” in interviews that the DIA detailees “conducted independent intelligence analysis resulting in analytic conclusions and products” (page 6). According to the contemporaneous written record, however, at least one of the DIA detailees said that “[alt no point did I prepare an intelligence estimate or publish anything I had written” during her involvement in the work under review. In any event, the terminology that individuals in informal interviews may have used or acquiesced to, advertently or inadvertently, cannot alter the nature of the work they actually did or did not do. In this case they did not produce or disseminate “intelligence assessments” or “Intelligence Analysis” on behalf of OUSD(P) or anyone else.

F. The Relevant Orders and Directives Describe Intelligence Roles and Activities, They Do Not Proscribe Policy Activities (U)

(U) The Report refers to definitions from DoD guidance dealing with intelligence agencies and intelligence activities. It endeavors to apply these definitions to policy activities undertaken for policy purposes within OSD. In so doing, the Draft Report transforms these definitions into restrictions on what policy offices may appropriately do.

(U) There is no authority to support the view that definitions describing the activities of intelligence agencies also apply to policy offices, or constitute limitations or prohibitions on the activities that policy offices may appropriately conduct. To demonstrate the fallacy of that thinking, one need only return to the relevant definitions.

(U) As discussed above, DoD Directive No. 5240.1 (Section 3.1) defines “Intelligence Activities” as:

“The collection, production, and dissemination of foreign intelligence and counterintelligence by DoD intelligence components authorized under [Executive Order 12333].”

(U) Executive Order 12333, “United States Intelligence Activities,” December 4,

1981, Section 3.4(e), defines “intelligence activities” as “all activities that agencies

within the Intelligence Community are authorized to conduct pursuant to this Order.”

Section 3.4(f) defines “Intelligence Community and agencies within the Intelligence

Community” as “the following agencies or organizations,” among which, as noted above,

OSD and OUSD(P) do not appear.

(U) DoD Directive No. 5240.1, Section 3.4, similarly defines “DoD intelligence components” as “[a]ll DoD Components conducting intelligence activities, including” a list of named DoD elements among which, again as noted above, OSD and OUSD(P) do not appear. But Section 2.1 of DoD Directive No. 5240.1 does define “DoD Components” to include the Office of the Secretary of Defense. Thus, as also noted above, the Directive distinguishes “all DoD Components” from “DoD Components conducting intelligence activities.” In consequence, the Directive’s Section 3.1 definition of “Intelligence Activities” by its terms only encompasses “DoD intelligence components,” not “all DoD Components,” as discussed above.

(U) The above definitions make two things clear about “Intelligence Activities”:

1. They constitute a process that entails collection, production “and” (not “or”) dissemination of foreign intelligence or counterintelligence, and

2. They are activities conducted by intelligence agencies, and not policy or other assessments or critiques by non-intelligence offices, even if these activities have similarities with “intelligence activities” performed by intelligence “agencies” or “components.”

(U) The Draft Report in effect expands the definition of “Intelligence Activities” contained in Directive 5240.1, Section 3.1, by dropping the restrictive clause “by DoD intelligence components authorized under [E. 0. 12333].” In other words, by asserting that OUSD(P) (admittedly not a “DoD intelligence component”) engaged in “Intelligence Activities,” the Draft Report obviously regards those activities as something that can be done by an entity that is not an “intelligence component.” The Draft Report thus appears to define “Intelligence Activities” as “the collection, production, and dissemination of foreign intelligence and counterintelligence” simply, regardless of by whom or what.

(U) This re-definition not only is incorrect on its face but in practice would lead to absurd results, as reference to the definition of “foreign intelligence” demonstrates. The term “foreign intelligence” appears in the definition of “Intelligence Activities,” i.e., the “collection, production, and dissemination of foreign intelligence and counterintelligence by DoD intelligence components authorized under” E.O. 12333. Both E.O. 12333 (Section 3.4(d)) and DoD Directive 5240.1 (Section 3.2) define “Foreign intelligence” as “information relating to the capabilities, intentions and activities of foreign powers,

organizations or persons, but not including counterintelligence except for information on international terrorist activities.”

(U) This definition of “foreign intelligence” is quite broad. The New York Times, for example, routinely engages in the collection (gathering and reporting), production (writing and editing) and dissemination (publication) of information relating to the “capabilities, intentions and activities of foreign powers, organizations or persons.” In the same vein, State Department Foreign Service officers, stationed both abroad and in Washington, constantly, through their contacts with foreign officials and others, learn about the “capabilities, intentions, and activities of foreign powers, organizations, or persons”; they report this information, which is used by the regional and other bureaus of the State Department to produce memoranda containing assessments and policy recommendations, which, in turn, are disseminated to officials throughout the government. Thus, if one were to accept the Draft Report’s modification of the definition of “intelligence activities,” one would have to conclude that the New York Times and State Department Foreign Service officers routinely engage in “intelligence activities.”

(U) Similarly, OUSD(P) routinely deals with “information relating to the capabilities, intentions, and activities of foreign powers, organizations, or persons.” For example:

• (U) Policy personnel routinely meet with foreign counterparts, at both the leadership and desk officer levels. These encounters occur at international meetings and conferences, formal defense bi-lateral consultations, and formal or informal one-on-one meetings. During such meetings, policy personnel acquire “foreign intelligence” information which is typically recorded in Memoranda for the Record, e-mails, etc.

• (U) In addition, policy personnel seek out other sources of information about “the capabilities, intentions, and activities of foreign powers, organizations, or persons,” for example, by attending academic or other conferences, or by talking to knowledgeable academics or other non-government experts on relevant subjects.

• (U) On the basis of this information and other sources (including “open source” intelligence, diplomatic reporting, as well as intelligence reports), Policy personnel prepare memoranda containing their analyses of foreign situations and associated policy recommendations. Almost all the work of regional offices, and much of the work of functional offices, deals with “the capabilities, intentions, and activities of foreign powers, organizations, or persons.”

• (U) These memoranda are disseminated within OUSD(P), to the Joint Staff and other DoD components, to the Defense Department leadership and to interagency colleagues.

(U) If this and similar activity were to be considered “Intelligence Activities,” then attempting to follow the Draft Report’s recommendation that “internal controls” be established to ensure that “Intelligence Activities” are not performed within OUSD(P) would be tantamount to shutting down OUSD(P) altogether.

(U) In fact, the guidance and authorities discussed here and in the Draft Report impose no restrictions on activities involving analyses, evaluations, assessments, critical reviews, or even alternative judgments by non-IC offices, not even if the subject of such analyses, etc. is intelligence reporting or intelligence products furnished by the IC, nor even if such analyses, etc. lead to judgments about intelligence information furnished by the IC that differ from the IC’s judgments about the same information.

(U) Where the relevant guidance intends to prohibit or regulate activities by non-IC offices, it does so in clear terms, and in only two instances: the prohibition on engaging or conspiring to engage in assassination (E.O. 12333, Section 2.11; DoD Directive No. 5240.1, Section 4.4); and the prohibition on all DoD Components from conducting or providing support for the conduct of special activities except as the Directive otherwise provides (DoD Directive No. 5240.1 Section 4.3). Other than these two cases, the relevant guidance does not proscribe any activities by non-IC offices. In particular it lacks any limitation on analyses or assessments by Policy offices of Intelligence Community information and products. There is no basis for characterizing the admittedly lawful and authorized work under review as “inappropriate.”

VI. OUSD(P) NONCONCURRENCE (U)

A. With the Findings of the Draft Report (U)

(U) For all the reasons stated in these comments, OUSD(P) does not concur in any finding expressed in the Draft Report except the finding that the activities reviewed were lawful and authorized, and specifically does not concur in incorrect assertions (e.g., at pages 4 and 14):

• (U) That OUSD(P) “developed, produced and then disseminated alternative intelligence assessments on the Iraq and al-Qaida relationship, which were inconsistent with the consensus of the Intelligence Community, to senior decisionmakers”;

• (U) That the actions reviewed were allegedly “OUSD(P)” activities;

• (U) That the actions reviewed were allegedly “inappropriate given that the products did not clearly show the variance with the consensus of the Intelligence Community and were, in some cases, shown as intelligence products”;

• (U) That there was an alleged “expanded role and mission of the OUSD(P) from Defense Policy formulation to alternative intelligence analysis and dissemination”;

• (U) That anything inappropriate occurred because “OUSD(P) lacked the management controls to ensure that Intelligence Activities were not performed, and that when Policy disagreed with the Intelligence Community, products produced by Policy clearly showed the variance with the Intelligence Community”;

• (LT) That OUSD(P) had a responsibility to, but “did not provide ‘the most accurate analysis of intelligence’ to senior decision-makers”; and

• (U) That any OUSD(P) activities, in response to requests by the Deputy Secretary, the Secretary of Defense or otherwise, constituted “Intelligence Activities.”

B. With the Recommendations of the Draft Report (U)

(U) For all the reasons stated in these comments, OUSD(P) does not concur in any recommendation expressed in the Draft Report, and specifically does not concur in the recommendations (page 14) that the Under Secretary of Defense for Policy:

“a. Establish internal controls so that ‘Intelligence Activities’ are not performed within the Office of the Under Secretary of Defense for Policy” — as OUSD(P) did not perform any “Intelligence Activities” and no such “internal controls” are needed.

“b. If in its policy formulation role, there is disagreement with the Intelligence Community consensus:

“(1.) Use existing procedures within the Intelligence Community to request an Alternative Judgment” — as existing IC procedures for producing “alternative judgments” do not apply to non-IC offices and are irrelevant to critiques by policy offices of IC work.

“(2.) Clearly articulate in policy products the Intelligence Community consensus and the basis for disagreement or variance from the Intelligence Community consensus” — as such a requirement would inappropriately constrain policy work by requiring policy offices to vet every policy product with the IC in order to determine whether or not it disagreed or varied with an IC “consensus” and — if it did -- to articulate the IC “consensus” in the policy product.

(U) Accordingly, OUSD(P) has taken no actions, and plans none, in response to the proposed recommendations.

VII. CONCLUSION (U)

(U) Bipartisan reports and studies by various commissions and congressional committees since the 9/11 attacks have stressed the need for hard questions and alternative thinking on the part of the Intelligence and Policy Communities alike. The motivation behind such observations has been a broadly held consensus that the Intelligence Community suffered major failures in its assessments of several key threats and issues before both the 9/11 attacks and the recent Iraq war. As the WMD Commission wrote, to quote just one such report:

“We conclude that good-faith efforts by intelligence consumers to understand the bases for analytic judgments ... are entirely legitimate. This is the case even if policymakers raise questions because they do not like the conclusions or are seeking evidence to support policy preferences. Those who must use intelligence are entitled to insist that they be fully informed as to both the evidence and the analysis.”⁵³

(U) The conclusions in the Draft Report reflect a disturbing departure from the trend in all these reports and studies to encourage the type of alternative thinking that motivated the work reviewed in this Project. By mischaracterizing that work as inappropriate “intelligence assessments,” the Draft Report fundamentally misinterprets what the work actually was — namely, a critical assessment by OSD, for policy purposes, of IC reporting and finished IC products on contacts between Iraq and al-Qaida. Such conclusions, if sustained, would have a dampening effect on future initiatives challenging intelligence assessments. The facts do not justify such conclusions.

(U) The work found “inappropriate” was an exercise in alternative thinking that the second most senior civilian in this Department directed his subordinates to prepare and brief to the most senior official of this Department. The latter, after receiving the draft briefing, directed that it be shared with the DCI. When the Deputy National Security Advisor requested the briefing, the Deputy Secretary’s office directed that it be given to him. These are the activities that the Draft Report characterizes as “inappropriate,” because it considers them to be “production” and “dissemination” of an “alternative intelligence assessment” contradicting assessments of the “chartered-intelligence community.” If the OIG actually believes that it was inappropriate for the Deputy Secretary of Defense to have some non-IC OSD staff members do a critical assessment of some IC work on a subject of major significance for national security, inappropriate for the Secretary of Defense to share the OSD work with the DCI, and inappropriate for the Deputy Secretary to share the work with the Deputy National Security Advisor when requested by the latter, the OIG should say so directly instead of

~‘ (U) Commission on the Intelligence Capabilities of the United States Regarding Weapons of Mass Destruction, Report to the President of the United States (31 March 2005), p. 189.

finding fault with subordinate OSD offices and staff members who did as they were instructed to do.

(U) The proposed recommendations would put a straightjacket on not only the type of work reviewed here but also the large majority of work routinely done in OSD, particularly in OUSD(P).

• (U) By having OUSD(P) to articulate the Intelligence Community consensus in any policy products that may vary from an IC “consensus” and the basis for such variance, the proposed recommendations would inappropriately constrain policy work. They would require policy offices to vet every policy recommendation or analysis with the IC in order to determine whether or not it disagreed or varied with an IC “consensus.” The proposed recommendations would also burden policy offices with a requirement to articulate the IC “consensus” when the IC itself should do so.

• (U) By having OUSD(P) to seek an “Alternative Judgment” from the IC whenever any OUSD(P) product disagreed with IC views, the proposed recommendations would seriously constrain and deter OUSD(P) personnel from articulating alternative views about the same information on which the IC’s assessments were based.

• (U) By mischaracterizing alternative reviews of IC work as “Intelligence Activities,” the conclusions of the Draft Report would chill the vigorous debate and hard questioning that most observers have recognized as necessary to avoid the types of intelligence failures experienced in the recent past.

(U) We strongly urge a reconsideration and major revision of the Draft Report and the conclusions expressed therein.

Eric S. Edelman

Under Secretary of Defense for Policy

APPENDIX A: DI POLICY NOS. 004 AND 005 604 0045 P 002

DEC-22-2006 11:35 DOD IG IWItL ~ ion

,~1Lcn1~1Uve AnalysES ~t~o1scy a~c

Directorate for Analysis and Production

Policy Statement

DI Policy No: 004- Date: 13 July 2001

Subject: &v~~O)1Cy

1. To remain relevant, intelligence analysis must pit clear conclusions that are actionable by the customer To be comprehensive, intelhgcnce analysts must bnng to bear the best collaborative effort of the broadest set of expertise These coinpetreg Lmpcrativea pose a dynamic, eontirnung challenge to intelligence analysis.

2. The normal process of coordination demands that analysts work with others possessing relevant expertise to develop consensus analysis. Often, this process produces compromises employing such tecimiques as numerical ranges (10-30 nussiles) or ambiguous word choices (possible, probable may could) Where compromises represent simple shadcs of thfference, tins outcome is acceptable where they mask genuine analytic alternatives, they do the customer a disservicc,

3. This alternative analysis policy encourages analysts to avoid those compromises wbi~h pass analytic ~inty along to the customer in the guise of authoritative analysis. An~ysts must cleariy.and

concisely-present what they know. what they do not know (and therefore assume), and then what they think (analysis). Genuine analytic differences, based on different evidence, assumptions, or methodologies, must be brought to the customer’s attention in a plain and succinct manner. Ow customers understand that our work is fraught with uncertainty, and they appreciate our riuanced attempts to explain what our differences axe, and why they occur. Analysts arc encouraged to resolve analytic differences by presenting alternative analysis within their products.

4. The need to promote sound altcrnativà analysis does not absolve the analyst from the require ant to collaborate. Rather, it frees the analyst from the need to resort to compromise just to reach a conclusion.

I//SIGNED/Il

CARYN A. WAGNER

Deputy Director for

Analysis and

Production

htp:fAw4ia.ic.gov/adminJdijpolicjc&pojjcy/~y~.~~,J 12/22/2006

DEC-22—2006 1t~35 DCD IS INTEL 703 604 0045 PlOt₂

Alternative Judgments t’ohcy dge U

Directorate for Analysis and Production

Policy Statement

DI Policy No: 005 - Date: 05 June, 2001

SUBJECT: ~ ‘iv4gjuqnts Policy

1. Curiosity and integrity are the hallmarks of good analysis. The best analysts constantly strive to explain the apparently inexplicable, while submitting their work to the rigorous scrutiny of their peers. The principal goal of intelligence analysis is to provide our customers with the most expert, fused₁and rnuhi-disciplinaryjudgements jiossible. Intellectual honesty and analytic rigor require processes that enable the analyst to present ideas and concepts which run counter to the prevailing wisdom or challenge the consensus. This policy memo establishes a process for promulgating alternative judvnents, consonant withthc tights and responsibilities of all analysts in their obligation to provide thebcst possible analysis.

2. The first and preferred method for incorporating an analytic alternative is through the standard process of coordination. Analysts arc expected to marshal their facts, build coherent arguments, and defend those arguments while coordinating with other experts across the Intelligence Commu* In the vast majority of cases, analytic judgments either stand or fall on the merits oftheir cvidentiary base, intrinsic logic and quality, In those rare Stances where analysts build a strong case, but cannot achieve consensus support for their analysis, an alternatIve judgment is justified.

3. Even with the existing venues for collaboration, there remain opportunities for the excLusion of analytic alternatives. Due to bureaucratic realities, the demands of time-sensitive taskings, or the unwillingness to accept contrary opinions, sound alternative judgments may be lost to the customer. To ensure there is an institutional cowl-of-last-resort for such analysis, this policy memo establishes the following procedure for raising alternative analytic judgments and giving them due consideration:

• Analysts who have presented analytic judgments through normal coordination, but whose judgments have been rebuffed, can produce an Alternative Judgment (Al) to inlbnn the senior leadership. The purpose of the Al is to provide visibility to the senior leadership of an alternative judgment Analysts who produce an Al do so secure in the knowledge that they have improved the prevailing analysis and reinforced the collaborativc process.

• The analyst has several responsibilities in the decision to produce an AS. First the analyst must permit the coordination process to play out an AS may net be used to circumvent normal coordination. Second, the Al must dispassionately (and succinctly) explain both the prevailing judgment and the alternative, without prejudice to the fonncr. Finally, the analyst must display trust in the senior leadership, both to give submitted £15 due consideration for further dissemination, and to protect the analyst’s best interests once they have produced an Al.

• The Al must adhere to the stated format (Enclosure I). This format forces the analyst to focus on the substance of the prevailing and alternative judgments, while avoiding the emotional or

http’i/www.dia.ic.gov/admin/di/poicies/poficy/oo5,htgnl 12/22/2006

DEC-22—2006 11:35 DOD IG INTEL 703 604 004b₁₁F~UOj~,

fl.ILCZJI4LIVC JuUg4UCflLS IUL1C~ age —

bureaucratic conditions which engendered the need for an Al in the first place. In the end, the Al is all about the analysis. Upon completing the Al, the analyst forwards it through the immediate Supervisor/Office Senior Intelligence Officer (510) to the Group SlOfkesearch Director (RD). The Supervisors/Office SIOs review the Al for format and completeness. The Group 510/RD reviews the AJ to ensure it accurately describes the competing analyses.

• The Group 510/Ri) then determines whether to: (I) present the Al to the Directorate’s Senior Analytic Review Board (SARB), or (2) return the & as not warranting fin’ther dissemination due to insufficient nou-responsiv; or unsubstantiated analysis. In the event the Group 510/RI) rejects the Al. he/she will return it to the analyst providing written explanation for the rejection; be/she will also provide a copy of the Al and the basis for the rejection to the Directorate Research Director (01-Ri)).

• The DX-RD will convene a panel of the Group SIOs/RDs to consider an Al selected for presentation by the appropriate Group 510/RD. The SAfl Will determine whether to (I) return the Al (due to insufficiency, as above) to the originating analyst, or (2) ratify the Al for inclusion in all appropriate products. The SARB wilt work in an expedited manner in order for the Al to be included ~n products already in draft. The SARB will recommend a means of dissemination to the DI, who will be the final authority on how an Al is promulgated,

4. The desired outcome of this alternative judgment process is to produce the highest quality intelligcrscc, while ensuring that critical alternative judgments are considered. The normal collaborative process should resolve most analytic differences well before the creation of an Al. The specific nature of the Al format is designed to provide discipline to the process, while providing the senior leadership sufficient information to mae an informed decision on the merits of the alternatives. The use of the SIOs/RDs as principal reviewers, and the oversight of the DI-RD, invigorate the senior analytic leadership in the process, while in no way replacing or circumventing the established role of the chainof-command. The analyst retains the right to submit alternative judgments, reinforced by the knowledge of an established process to ensure fundamental fairness. All of these intended outcomes combine toward the ultimate goal of encouraging analytic rigor.

f/ISICJNED///

CARYN A. WAGNER

Deputy Director for

Analysis and

Production

~ 12/22/2006

APPENDIX B: COMMENTS ON OIG’S ANSWERS TO SENATOR LEVIN

(U) Q. 1. As explained in our comments, it is incorrect to attribute the briefing in question to “the OUSD(P).” It is also incorrect to characterize it as an “intelligence analysis.”

(U) Q. 2. The Draft Report ignores the October 2002 letter from DCI Tenet to Chairman Graham of the Senate Select Committee on Intelligence and other DCI statements to Congress, as discussed in our comments. The Draft Report hence cannot, and should not purport to, judge the extent to which the IC views as expressed in that letter (which says, inter alia, that “we have solid reporting of senior level contacts between Iraq and alQa’ida going back a decade” and “credible information indicates that Iraq and al-Qa’ida have discussed safe haven and reciprocal non-aggression”) were or were not compatible with the view that there was a “mature, symbiotic” relationship between Iraq and alQa’ida.” (Note that the briefing speaks of a “mature, symbiotic relationship” and not of the “mature, symbiotic’ cooperation” attributed to it in this answer (emphasis supplied).)

• (U) It is misleading to say, in the second paragraph of this answer, that the CIA “later dismissed the alleged” Atta meeting, as if the CIA’s later view rather than its contemporaneous view is relevant to this question. During the relevant period in 2002, the CIA never went so far as to “dismiss” the alleged meeting.

• (U) The final sentence of the second paragraph of this answer (“Within the OUSD(P), however, the different conclusions [i.e., the alleged “higher degree of cooperation”] were ‘entirely favored’ over the Intelligence Community’s views”) has no basis within the Draft Report or otherwise.

• (U) It is misleading, in the third paragraph of this answer, to quote the August 2002 CIA report stating that the CIA “could not document any joint operational activity between Iraq and al.-Qaida.” None of the work under review asserted that there had been any such activity.

(U) Q. 3. The Draft Report contains no analysis of the “underlying intelligence.” Thus, the assertion that the “alternative intelligence analysis that OUSD(P) produced” was only partially supported by it is itself not supported.

• (U) There is no basis for asserting that the view that there was a “mature, symbiotic relationship” between Iraq and al-Qa’ida “was based primarily on the alleged 8-9 April 2001 meeting in Prague between Mohammed Atta and alAni.” In fact, that view was based on a series of intelligence reports.

• ( ) The Draft Report’s answer to this question misleadingly brings in the purported views of the Czech intelligence service in ,which can have no bearing on whether statements made in 2002 were or were not supported by the (then available) underlying intelligence.

(U) Q. 6. It is misleading to describe the briefing to the Deputy Assistant to the President for National Security Affairs (which the Vice President’s Chief of Staff attended, at least in part) as a “briefing to the Office of the Vice President.” It is tendentious to describe the Atta slide as “previously unseen,” as the slide did not previously exist. It is incorrect to assert, as this answer does, that this new slide presented the alleged Atta meeting “as fact” (page 27). Nowhere does the slide describe the meeting as “fact.” To the contrary, the slide repeatedly uses phrases such as “Czech service reports that Atta visited ...,“ “despite press reports of conflicting information, Czech Interior Minister ... stands by previous Czech ... reporting,” “Atta reportedly held meetings...,” and “Atta reportedly arrives in Prague....”

(U) Q. 7. There is no evidence that the authors of the Draft Report reviewed the available intelligence on the relevant issues. Thus, it would appear that they are not in a position to assert that the briefing in question was or was not supported by it.

• (U) It is incorrect to say that the differences between the three versions “altered the overall message presented to each audience.” The Draft Report does not discuss the “overall message” of each version of the briefing and does not analyze how the “overall message” of one version relates to the “overall message” of another version.

(U) Q. 8. The Draft Report endorses the questioner’s view that the “fundamental problems” slide “undercut” the IC. The Draft Report provides no evidence that the IC was in fact “undercut” or harmed in any way⁵⁴, no explanation how this would have happened, or what effects it might have had, if any. The implication is that the IC may not be criticized at all.

(U) Q. 9. The Draft Report’s affirmative answer to this question is not supported by the evidence provided, which nowhere reviews the “available intelligence.”

• The Draft Report does not compare the briefing’s statements with the statements by DCI Tenet in his October 7, 2002, letter to Senator Graham and other statements to Congress. For example, DCI Tenet said that “We have solid reporting of senior level contacts between Iraq and al-Qaida going back a decade.”

~“ (U) The relevant dictionary definition of “undercut” is “to undermine or destroy the force, value or effectiveness of’ (Webster’s Ninth New Collegiate Dictionary).

(U) Q. 10. The last sentence of the first paragraph of this answer (“The CIA was not given advance notice or an opportunity to respond to the critique because the OSD considered it an internal OSD product”) implies that somehow the OSD view was questionable or incorrect. However, the product indisputably was an internal OSD product, and there is no reason why CIA should have been informed of it, any more than OUSD(P) was informed of the DIA memos of August 9 and 14, 2002, discussed on page 9 of this Draft Report.