Martin Indyk, with whom I served on the NSC staff early in the Clinton administration and who went on to be a.s.sistant secretary of state for Near Eastern Affairs, ill.u.s.trated for me the mind-set that we were all operating with in the mid-1990s. Martin and I were convinced that Iraq had weapons of ma.s.s destruction because Saddam had an entire organization dedicated to concealing them. "We observed how they operated," Martin said. "Saddam refused to account for the material that was missing from the previous war, and logically it did not make sense, since if he would just come clean he could get out of sanctions and we would be screwed."

"I remember going to bed at night," Martin recalled, "expecting to wake up the next morning and find that UNSCOM would go to the secret site and catch Saddam red-handed. We'd wake up in the morning and there was nothing there. There was never anything there. With the benefit of hindsight, we should have thought: wait a minute, if we never find it, maybe it is not there. I didn't think about the possibility that Saddam was bluffing us."

I did not believe he was bluffing, either. With the quality of UN inspections growing weaker over time, the political will to maintain sanctions fading, and Saddam's coffers ballooning through the Oil-for-Food program, I had little doubt in my own mind what Saddam was up to. I believed he had WMD, and I said so.

From then on, after UNSCOM's departure, we had to rely more on a.n.a.lysis and extrapolation of more nuanced technical data. We divorced technical a.n.a.lysis from our understanding of Iraqi culture, however, and this hurt us in central ways. We failed, for example, to factor in how the regime's harsh treatment of its citizens would make truthful reporting to superiors on the status of weapons programs less likely. We did not fully consider the impact of nearly a decade of international sanctions, UNSCOM inspections, continuous overflights, and U.S. military actions. Yet Saddam gave us little reason to believe that he had changed his stripes or his trajectory.

Nevertheless, in 2002, to conclude that Saddam was not pursuing WMD, our a.n.a.lysts would have had to ignore years and years of intelligence that pointed in the direction of active programs and continuing evidence of aggressive attempts on Iraq's part to conceal its activities. Even with more time, could a.n.a.lysts have concluded that Saddam had no weapons programs, or even the ability to quickly surge to produce the weapons themselves? I doubt it.

In retrospect, we got it wrong partly because the truth was so implausible. We knew plenty of countries that were working on WMD programs and desperately trying to conceal that fact. But we had no previous experience with a country that did not possess such weapons but pretended that it did. Saddam made a speech in June 2000 in which he said you cannot expect Iraq to give up the rifle and live only with a sword when his neighbor [Iran] had a rifle. After his capture in December 2003, Saddam was asked by George Pirro, an FBI Special Agent, what he had meant by that statement. Saddam said that he had two audiences in mind. One was the UN Security Council, as Saddam wanted the disarming of Iraq to be part of a broader disarming of the Middle East. The other audience was Iran. Saddam then said, "You guys just don't understand. This is a rough neighborhood."

There is another factor that few people outside the intelligence community would recognize or credit, and that is how the remedy for one so-called intelligence failure can help set the stage for another. Following the controversy over some of our missile a.n.a.lysis in the mid-1990s, a commission headed by Donald Rumsfeld had taken us to task for not leaning forward more boldly and imaginatively in projecting missile development in countries such as Iran and North Korea. In response, we began to give more weight in these a.s.sessments to what could could occur, rather than stopping with what we confidently knew. This is perhaps another way of saying, connect all the dots in order to warn adequately. I have often wondered if this was the prevailing sentiment among a.n.a.lysts as we did our Iraq work. Did it push us to be more a.s.sertive than we should have been? occur, rather than stopping with what we confidently knew. This is perhaps another way of saying, connect all the dots in order to warn adequately. I have often wondered if this was the prevailing sentiment among a.n.a.lysts as we did our Iraq work. Did it push us to be more a.s.sertive than we should have been?

Saddam was a genius at what the intelligence community calls "denial and deception"-leading us to believe things that weren't true. But he was a fool for not understanding, especially after 9/11, that the United States was not going to risk underestimating his WMD capabilities as we had done once before. The irony is that he could have allowed UN inspectors free run of the country-and if they found nothing, UN sanctions would have melted. In that case, he might be alive and living in a palace today. Without sanctions, he would be well on his way to possessing WMD. Before the war, we didn't understand that he he was bluffing, and he didn't understand that was bluffing, and he didn't understand that we were not we were not.

When we finally did complete the nineteen-day Estimate late on the evening of October 1, the doc.u.ment was rushed to Capitol Hill with the ink still wet on its covers.

The morning of October 2, 2002, twelve hours after we had delivered the NIE to Congress, the Senate Select Committee on Intelligence held a closed-door hearing to discuss its contents. My staff had informed the committee several days earlier that I would be unable to attend because I was required to be at the White House at the same time, ironically to meet with other congressional leaders. In my place, I sent John McLaughlin and Bob Walpole, the lead national intelligence officer for the NIE, to brief members. The members, though, seemed to have forgotten that I had advised them I could not be present. Several were upset about my absence and about the NIE having been delivered so late the night before, around 10:30 P.M. P.M. Their anger was misplaced-McLaughlin and Walpole were well qualified to respond to the senators' interests. Their anger was misplaced-McLaughlin and Walpole were well qualified to respond to the senators' interests.

With great difficulty, McLaughlin persuaded the committee to go ahead with the hearing, and then only by promising that he and I would return to brief the senators again two days later. We did that on Friday, October 4. That closed-door session was very contentious.

One senator asked us how our views differed from those of our British allies, who had just published their own white paper days before. Bob Walpole cited two points of divergence. First, he said, we differed by a few months with the British on how quickly Saddam could make a nuclear weapon. Second, we differed with the Brits on intelligence suggesting that Saddam had been trying to obtain uranium from Africa. Senator Kyl pointed out that there was reference to yellowcake in our Estimate. Walpole said, yes, we mention it as a possibility, but only after we say that we are much more worried about the 550 tons of yellowcake that Saddam already has access to inside Iraq. Even then, Walpole pointed out, yellowcake is not mentioned in the Key Judgments or in our uncla.s.sified paper.

As soon as we delivered the cla.s.sified Estimate to the Hill, calls began for us to instantly produce an uncla.s.sified version. This, too, was virtually impossible in the time allotted, but our efforts to be accommodating led to another major error. Someone came up with the bright idea of taking an uncla.s.sified white paper that the NIC had drafted months before on the same subject, and had sat unpublished on a shelf, and modifying it for this purpose. Doing so would be far faster than trying to come up with an uncla.s.sified version of the NIE. But there's a saying that "if you want it bad, you get it bad," and that was precisely what we got.

In an effort to meld the white paper and the NIE, a.n.a.lysts took the Key Judgments from the NIE, decla.s.sified them, and stuck them on the front of the white paper. Because they are written from the point of view of the entire intelligence community, NIEs are replete with statements such as "we a.s.sess that" and "we judge that." The white paper had been crafted in a different style, and in merging the two doc.u.ments, those responsible opted for the latter style. Out went the "we"s, and what remained were bolder a.s.sertions, such as "Saddam has." The cla.s.sified NIE already had too few cautionary "we judge"s in the Key Judgment section. Now, with a few strokes of a keyboard, the uncla.s.sified paper-the only one most Americans would ever see-came out sounding far too a.s.sertive, even though it did note that there were differences among specialists over issues such as the aluminum tubes and UAVs. The moral to the story is that white papers should never be written before a cla.s.sified estimate has been completed.

Following McLaughlin and Walpole's October 2 appearance before the Senate Select Committee on Intelligence, several Democratic senators demanded that a few sentences from the testimony be decla.s.sified and cleared for public release. The senators also wanted released some language that was contained in the cla.s.sified NIE but not in the uncla.s.sified white paper.

On October 7, McLaughlin signed a letter to them on my behalf containing the words they were seeking from the NIE:

Baghdad for now appears to be drawing a line short of conducting terrorist attacks with conventional or C.B.W. [chemical and biological weapons] against the United States.Should Saddam conclude that a U.S.-led attack could no longer be deterred, he probably would become much less constrained in adopting terrorist actions. Such terrorism might involve conventional means, as with Iraq's unsuccessful attempt at a terrorist offensive in 1991, or C.B.W.Saddam might decide that the extreme step of a.s.sisting Islamic terrorists in conducting a W.M.D. attack against the United States would be his last chance to exact vengeance by taking a large number of victims with him.

The letter also authorized the release of some dialogue between Senator Carl Levin and John McLaughlin, who testified in a closed hearing. The witness said that the likelihood of Saddam's initiating a terrorist attack in the foreseeable future was, in our judgment, "low," but that if Saddam felt cornered, the chances of his using WMD were "pretty high."

Democratic members of the committee released the letter to the media almost immediately, provoking a flurry of stories. The articles suggested that the letter contradicted President Bush's a.s.sertion on the imminent threat posed by Iraq and implied that the use of force by the United States would only increase the likelihood that Saddam would either use WMD himself or share it with terrorists. The articles prompted a frantic call from Condi Rice. She wanted me to "clarify" the issue right away. So, at her request, I spoke with a New York Times New York Times reporter who was working on the story. "There was no inconsistency in the views in the letter and those of the president," I told the reporter. The sentence seized upon in the letter was about a judgment call as to whether and when Saddam might use WMD and whether he might share them with terrorist organizations. We labeled our views as "low confidence" judgments-in other words, we were not very sure we had a good idea what Saddam would do if cornered. reporter who was working on the story. "There was no inconsistency in the views in the letter and those of the president," I told the reporter. The sentence seized upon in the letter was about a judgment call as to whether and when Saddam might use WMD and whether he might share them with terrorist organizations. We labeled our views as "low confidence" judgments-in other words, we were not very sure we had a good idea what Saddam would do if cornered.

In retrospect, I shouldn't have talked to the New York Times New York Times reporter at Condi's request. By making public comments in the middle of a contentious political debate, I gave the impression that I was becoming a partisan player. That certainly wasn't my intention. reporter at Condi's request. By making public comments in the middle of a contentious political debate, I gave the impression that I was becoming a partisan player. That certainly wasn't my intention.

The intelligence reports and a.n.a.lysis used over the years on the WMD issue, and repeated in the NIE, were flawed, but the intelligence process was not disingenuous nor was it influenced by politics. Intelligence professionals did not try to tell policy makers what they wanted to hear, nor did the policy makers lean on us to influence outcomes. The consistency of our views on these weapons programs was carried forward to two presidents of different political parties who pursued vastly different courses of action. Even though the daily reports the president saw in the run-up to the production of the NIE were uneven and a.s.sertive in tone, and at times more a.s.sertive on some issues than the NIE, they were a reflection of honest a.n.a.lysis.

Policy makers also have the responsibility to challenge the a.n.a.lysis they receive. Their uncritical att.i.tude in this case was highlighted by a question posed by Brent Scowcroft in a recent speech: "What happens when the intelligence community provides intelligence that policy makers want to hear?" He could have added: particularly when war and peace hang in the balance. particularly when war and peace hang in the balance.

An NIE had never been relied upon as a basis for going to war, and, in my view, the decision to invade Iraq was not solely predicated on this one. But if we had done a better job in all our a.n.a.lysis and in this NIE, war critics would have had a harder time today implying that "the intelligence community made us do it."

The notion that we somehow cooked the books on the Iraq NIE is only part of current mythology. Maybe the greater exaggeration is the profound effect the NIE supposedly had on decision makers. In a little-remembered article in April 2004, the Washington Post Washington Post reported, "No more than six senators and a handful of House members who did not serve on the house and senate Intelligence Committees read beyond the five-page National Intelligence Estimate executive summary, according to several congressional aides responsible for safeguarding the cla.s.sified material." The full NIE ran some ninety pages. reported, "No more than six senators and a handful of House members who did not serve on the house and senate Intelligence Committees read beyond the five-page National Intelligence Estimate executive summary, according to several congressional aides responsible for safeguarding the cla.s.sified material." The full NIE ran some ninety pages.

Some who later rightly criticized the NIE had previously made their own public statements that went beyond what was in the Estimate. Senator Jay Rockefeller, the respected ranking Democrat on the Senate Intelligence Committee, said on the floor of the Senate on October 10, 2002, that "There is unmistakable evidence that Saddam Hussein is working aggressively to develop nuclear weapons and will likely will likely have nuclear weapons within the have nuclear weapons within the next five years next five years" (emphasis added). The first sentence of the Key Judgments of the October 2002 estimate itself says only that "if left unchecked," "if left unchecked," Saddam Saddam "probably" "probably" will have a nuclear weapon will have a nuclear weapon "during this decade" "during this decade" (emphasis added). (emphasis added).

Rather than being "unmistakable," the evidence was a matter of some dispute among a.n.a.lysts, a point made clear in pages of dissenting opinions in the NIE. Rockefeller went on to remind his colleagues of the same history that caused our a.n.a.lysts much concern. He said, "We also should remember we have always underestimated the progress Saddam has made in development of weapons of ma.s.s destruction."

Congress was not alone in its lack of genuine interest in the NIE before the war. Senior administration officials in the NSC, Department of Defense, and elsewhere had also put the doc.u.ment at the bottom of their reading lists. Everyone seemed to think they knew either what was in the doc.u.ment or what ought to be in it.

Few people may have read the NIE, but in no way does this excuse the many shortcomings of our Iraqi a.n.a.lysis over the years, in the Estimate or in the testimony we presented to Congress. Misinformation and misimpressions go to the heart of our credibility, our mission, even our reason for being.

Given what we knew then, the NIE should have said:

We judge that Saddam continues his efforts to rebuild weapons programs, that, once sanctions are lifted, he probably will confront the United States with chemical, biological and nuclear weapons within a matter of months and years. Today, while we have little direct evidence of weapons stockpiles, Saddam has the ability to quickly surge to produce chemical and biological weapons and he has the means to deliver them.

We should have said, in effect, that the intelligence was not sufficient to prove beyond a reasonable doubt that Saddam had WMD. The evidence was good enough to win a conviction in a civil suit but not in a criminal case. Would we have gone to war with such conclusions? I don't believe the war was solely about about WMD, so probably yes. But more accurate and nuanced findings would have made for a more vigorous debate-and would have served the country better. WMD, so probably yes. But more accurate and nuanced findings would have made for a more vigorous debate-and would have served the country better.

In the spring of 2004, during one of my final appearances before the House Permanent Select Committee on Intelligence, Congressman Norm d.i.c.ks commented on the NIE. Norm is a longtime friend of the intelligence community and of mine personally, yet he had harsh words that day. Regarding the Estimate, and the faith he had in me, he said, "We depended on you, and you let us down." For me, it was one of the lowest moments of my seven-year tenure, because I knew he was right.

CHAPTER 18

No Authority, Direction, or Control

Mr. President," I said one morning in March 2003, "the vice president wants to make a speech about Iraq and al-Qa'ida that goes way beyond what the intelligence shows. We cannot support the speech, and it should not be given."

The Iraq WMD issue had been around for years. People believed they knew it backward and forward. There was no raging debate within the administration about our conclusions. But there was was debate, intense focus, and, in the eyes of some a.n.a.lysts, pressure regarding the question of Iraq's relationship with al-Qa'ida and complicity in 9/11. We could go as far as outlining contacts between Iraq and al-Qa'ida going back a decade, to Bin Ladin's time in the Sudan, to Abu Musab al-Zarqawi finding safe haven in Iraq, and to at least a dozen Egyptian Islamic Jihad operatives who showed up in Baghdad in the spring and summer of 2002. We could cite training that may have been provided, particularly regarding chemical and biological weapons. But one thing is certain, we consistently told the Congress and the administration that the intelligence did not show any Iraqi authority, direction, or control over any of the many specific terrorist acts carried out by al-Qa'ida. debate, intense focus, and, in the eyes of some a.n.a.lysts, pressure regarding the question of Iraq's relationship with al-Qa'ida and complicity in 9/11. We could go as far as outlining contacts between Iraq and al-Qa'ida going back a decade, to Bin Ladin's time in the Sudan, to Abu Musab al-Zarqawi finding safe haven in Iraq, and to at least a dozen Egyptian Islamic Jihad operatives who showed up in Baghdad in the spring and summer of 2002. We could cite training that may have been provided, particularly regarding chemical and biological weapons. But one thing is certain, we consistently told the Congress and the administration that the intelligence did not show any Iraqi authority, direction, or control over any of the many specific terrorist acts carried out by al-Qa'ida.

Let me say it again: CIA found absolutely no linkage between Saddam and 9/11. At best, all the data in our possession suggested a plausible scenario where the "enemy of my enemy might be my friend," that is, two enemies trying to determine how best to take advantage of each other. In the world of terrorism, nothing is ever very clear, and the murkiness of the intelligence required an exhaustive effort to run down every lead to satisfy ourselves that there was no state complicity with al-Qa'ida's actions on 9/11.

We told the president what we did on Iraq WMD because we believed it. However, we did not bend to pressure when it came to a possible past Iraqal-Qa'ida connection. The absence of such a connection would have been impossible for others to disprove following an invasion, unlike WMD, which were either there or not. Those who say that we cooked the books or knowingly let the administration say things that we knew to be untrue are just wrong.

People often forget what it was like after 9/11. A senior a.n.a.lyst put it this way, "Intelligence is central to the Bush administration. Every single day it was the discipline around which they started their day. And then after 9/11, the first attack on American soil of any magnitude in sixty years, they were in fear. In fairness to them, people do not understand how G.o.dd.a.m.n dangerous we thought it was. The absence of solid information on additional threats was terrifying."

It took us a while to understand how important the Iraq connection was for some in the administration, but we learned quickly. The vice president and others pushed us hard on this issue, and our answers never satisfied him or some of our other regular "customers." Paul Wolfowitz and Scooter Libby, for example, were relentless in asking us to check, recheck, and rerecheck. Wolfowitz's strong views on the matter were no secret. He even wrote a blurb for Laurie Mylroie's 2000 book, Study of Revenge: Saddam Hussein's Unfinished War Against America Study of Revenge: Saddam Hussein's Unfinished War Against America, in which he said the book "argues powerfully" that the perpetrator of the 1993 World Trade Center attack was actually "an agent of Iraqi intelligence," and it asks, if that is true, what that would tell us about Saddam's ultimate ambitions.

The truth was that CIA was not initially prepared for the intense focus that the administration put on the Iraqal-Qa'ida relationship. We had devoted little a.n.a.lytic attention to it prior to September 11. We were instead consumed with the very hot war with Sunni extremists all over the world. People were coming to kill us. We had no preconceived conclusions on the Iraqal-Qa'ida connection-unlike our certainty on Iraqi WMD-and it would require us to start from the bottom up, do a zero-based review and look at the whole issue dispa.s.sionately. On one level, this was a blessing.

It was also a curse, because initially, and for some time, our answers to the elaborate, nuanced, and voluminous questions the administration asked were inconsistent and incomplete, and often had to be revisited. Early on we probably did not inspire much confidence in policy makers who knew their brief and knew where they wanted to end up. Senator Fritz Hollings once said that going to a press conference with Vice President Hubert Humphrey was like jumping into a swimming pool with Olympic champion Mark Spitz. Well, that was what it was like briefing d.i.c.k Cheney, Scooter Libby, and Paul Wolfowitz on this subject. They were smart, tough, and had command of the data. Initially, we did not. But over time, that changed in a dramatic way.

The first time I recall a briefing at our headquarters on Iraq and al-Qa'ida was in September of 2002. The briefing was a disaster. Libby and the vice president arrived with such detailed knowledge on people, sources, and timelines that the senior CIA a.n.a.lytic manager doing the briefing that day simply could not compete. We weren't ready for this discussion. We determined that from that moment on we would have multiple lower-level subject-matter expert a.n.a.lysts-people who knew a lot about a narrow range of topics-meet with them.

By November of 2002, we were ready for another visit from the vice president and his team. There was extensive preparation, practice sessions called "murder boards," and total collaboration between regional and terrorist a.n.a.lysts. The November meeting was described by a partic.i.p.ant this way:

Scooter Libby approached it like an artful attorney. An a.n.a.lyst would make a point and Libby would say, okay this is what you say. But there are these other things happening. So if this were true, would it change your judgment? And the a.n.a.lyst would say, well if that was true, it might. And Libby would say, well if that's true, what about this? And six "if that were trues" later, I finally had to stop him and say, "Yes, there are other bits and pieces out there. We've looked at these bits and pieces in terms of the whole. And the whole just does not take us as far as you believe. And everything else is just speculation. That was a push by policy makers to see how far we would go."

Some a.n.a.lysts viewed this kind of grilling as being pressured, but most did not. Their view was, if a country is about to go to war, policy makers are going to ask tough questions to understand all the elements of the issue. One senior a.n.a.lyst said to me, "Were they trying to push us and drive us? Absolutely. By the questions they asked and by the way they asked the questions again and again with changed nuances. They were trying to pull out every last iota of what we might say that supported where they wanted to go. But they are policy makers. It is our job to consider what they say, think about it, and write what we think. We stuck to our guns."

The truth is we were not ready to take a position on the Iraqal-Qa'ida question because there were differing views within the Agency about how to think about the issue. The division was between a.n.a.lysts who focused on specific regions and those who specialized in terrorism. This uncertainty was played out earlier in the year on June 21, 2002, when we produced the paper "Iraq & al-Qa'ida: Interpreting a Murky Relationship." In contrast with every other type of a.n.a.lysis, because of the nature of the threat, terrorism a.n.a.lysis by design takes weaker information and makes more aggressive conclusions, sometimes from information that regional a.n.a.lysts might discard.

The "Murky Relationship" paper was an academic exercise. Its "scope note" at the beginning explained that the paper was an effort to see what our conclusions might be if the most forward leaning explanations of our intelligence turned out to be true. The note read in part: "This intelligence a.s.sessment responds to senior policy maker interest in a comprehensive a.s.sessment of Iraqi regime links to al-Qa'ida. Our approach is purposefully aggressive in seeking to draw connections purposefully aggressive in seeking to draw connections, on the a.s.sumption that any indication of a relationship between these two hostile elements could carry great dangers to the United States" (emphasis added).

Regional a.n.a.lysts who focus on geographic areas believed that fundamental distrust stemming from stark ideological differences between Saddam and Usama bin Ladin, and the potential fear that Islamic extremism posed for Iraq, significantly limited the cooperation that was suggested by the reporting. The terrorism a.n.a.lysts who specialized in the broad range of terrorism and who wrote the paper took note of the ideological differences but believed to be credible the reporting that suggested a deeper relationship. The paper made clear that there were no conclusive signs between Iraq and al-Qa'ida with regard to terrorist operations. Yet it posited that there were enough data with regard to safe haven, training, and contacts to at least require us to be very concerned. Jami Miscik, our chief a.n.a.lyst, believed that the a.n.a.lysis should be published because of the risks to the United States, and it was.

In our shop, many saw this as almost too aggressive. Some a.n.a.lysts involved complained informally to an ombudsman, whom we had earlier appointed to deal with claims of politicization, that we had gone too far in coming up with our "murky" conclusion. As described to me by a senior a.n.a.lyst, "Barry [the ombudsman] sat us down and said: 'Grow up. This is not politicization. This is misunderstanding and hurt feelings.' The two groups need to sit down and hash it out."

Despite the fact that some of our a.n.a.lysts felt we had gone too far, many in the administration, such as Paul Wolfowitz and Scooter Libby, believed that the "Murky Relationship" paper had not gone far enough. Within a couple months the cla.s.sified doc.u.ment was being mocked in a Jim Hoagland column in the Washington Post. Washington Post. Hoagland's piece led with the rhetorical slap: "Imagine that Saddam Hussein has been offering terrorist training and other lethal support to Osama bin Ladin's al-Qa'ida for years. You can't imagine that? Sign up over there. You can be a Middle East a.n.a.lyst for the Central Intelligence Agency." Hoagland's piece led with the rhetorical slap: "Imagine that Saddam Hussein has been offering terrorist training and other lethal support to Osama bin Ladin's al-Qa'ida for years. You can't imagine that? Sign up over there. You can be a Middle East a.n.a.lyst for the Central Intelligence Agency."

Other senior administration officials questioned our preliminary a.n.a.lysis. The Senate Intelligence Committee later uncovered an internal Pentagon memo sent to both Paul Wolfowitz and Doug Feith saying that although the facts represented in the "Murky Relationship" paper were good, CIA's a.n.a.lysis attempted to "discredit, dismiss, or downgrade" much of the reporting, and our interpretations "should be ignored."

During the late summer of 2002, we started working on a more comprehensive paper that would explain what we knew and suspected about Iraq's involvement with terrorism. While we could not make the al-Qa'ida connection, there was no doubt that Saddam was making large donations to the families of Palestinian suicide bombers and was known to be harboring several prominent terrorists, including Abu Nidal, a ruthless killer responsible for attacks on El Al ticket counters in Rome and Vienna in 1985, resulting in 18 deaths and injury to 120 people. Saddam also gave refuge to one of the individuals still being sought for the first World Trade Center bombing.

We were still gathering material for the comprehensive paper when we received an offer from a Pentagon group working under Doug Feith to share with us their observations on the case for a connection between Iraq and terrorism. Although the suggestion was a bit odd-since it was coming from people in the policy shop, not people in intelligence positions-we agreed to hear them out. A small group of Pentagon officials showed up at CIA headquarters on August 15, 2002.

Present from the Pentagon were Feith; Richard Haver, a longtime civilian intelligence professional who had worked for d.i.c.k Cheney in the first Bush administration; Vice Admiral Jake Jacoby, the head of the Defense Intelligence Agency; and several others from Feith's shop. Haver had been the one who dropped by my office in late December 2000 and hinted broadly that I was soon to be replaced by Don Rumsfeld. That had not been his only off-the-mark a.s.sessment. Shortly before September 11, 2001, he gave a speech at the National Security Agency during which he told the audience that the intelligence community was spending far too much time on terrorism.

Attending the meeting on our side with me were Ben Bonk, the deputy chief of our Counterterrorism Center; several a.n.a.lysts from Ben's staff; and a number of a.n.a.lysts from CIA's Directorate of Intelligence who were coordinating the forthcoming Iraq terrorism paper.

Feith's team, it turned out, had been sifting through raw intelligence and wanted to brief us on things they thought we had missed. Trouble was, while they seemed to like playing the role of a.n.a.lysts, they showed none of the professional skills or discipline required. Feith and company would find little nuggets that supported their beliefs and seize upon them, never understanding that there might be a larger picture they were missing. Isolated data points became so important to them that they would never look at the thousands of other data points that might convey an opposite story.

Tina Shelton, a naval reservist on Feith's team, gave the presentation, which was t.i.tled "Iraq and al-Qa'ida-Making the Case." She started out by saying that there should be "no more debate" on the Iraqal-Qa'ida relationship. "It is an open-and-shut case," she said. "No further a.n.a.lysis is required." This statement instantly got my attention. I knew we had trouble on our hands.

The briefing slides she used were equally self-certain. One slide said that Iraq and al-Qa'ida had a "mature, symbiotic relationship." Wrong. There was nothing in the intelligence to suggest a mature, symbiotic relationship. Another slide said there were "some indications of possible Iraq coordination with Al Qaeda specifically related to 9/11." By this point, the "Atta in Prague" story, which CIA had brought forward after 9/11, was eroding.

I listened for a few more minutes, trying to be polite, before saying, "That's very interesting." This was one of my rare moments of trying to be subtle. What I was really thinking was, This is complete c.r.a.p, and I want this to end right now.

Shortly thereafter I excused myself and pulled Jacoby aside. As an active-duty vice admiral and head of DIA, he worked for both Don Rumsfeld and me. Reverting to my normal blunt self, I told him, "This is entirely inappropriate. You get this back in intelligence channels. I want a.n.a.lysts talking to a.n.a.lysts, not people with agendas." Pentagon officials were later quoted anonymously in the media describing the same meeting but claiming that "the scales fell away" from the CIA's eyes when we saw their fine a.n.a.lysis. In fact, their a.n.a.lysis had little, if any, impact on us whatsoever.

Only much later did we learn that "Team Feith" had been going around briefing officials at the White House, the NSC, and the Office of the Vice President with a story similar to the one we found so weak in August. In these briefings they added an extra slide t.i.tled "Fundamental Problems with How Intelligence Community Is a.s.sessing Information." The slide went on to complain that we were being too picky and applying a standard of proof that we would not normally require. But we weren't too impressed with their work, either, especially their willingness to blindly accept information that confirmed preconceived notions. We came to call their approach "Feith-based a.n.a.lysis."

When the Pentagon's inspector general issued a report in February 2007 calling some of Feith's efforts "inappropriate" (which to my mind is the kindest thing you could say about it), Feith shot back. He said peddling his alternative intelligence was simply an exercise in "good government." Nonsense. This was an example of bad government. Policy makers are ent.i.tled to their own opinions-but not to their own set of facts. Feith's charts mischaracterized the intelligence. If policy makers want to reach their own judgments they can do so, so long as they say, "The views I am about to express are not supported by the DCI and his a.n.a.lysts." And Feith should have had the courage to tell us that his opening slide, shown to the White House, said in essence that CIA a.n.a.lysis stinks.

Our second paper on Iraq and al-Qa'ida, published in September 2002, was shared with only a small number of senior officials. As often happens, shortly after that report went out, new intelligence came in suggesting that there might have been greater contact regarding training between Iraq and al-Qa'ida. In light of that, we set out to vet and add these new details for an upgraded report that would be given wider dissemination among administration and congressional officials than the first, closely held doc.u.ment. Agency a.n.a.lysts went so far as to show a draft to Feith's team and to ask if they had any comments or objections to it. Feith's staffers said they did "but would make their views known through other channels." In retrospect this was a pretty clear warning that we were being second-guessed and undermined.

By December the revision was done, and we t.i.tled the report, "Iraqi Support of Terrorism." I asked, out of courtesy, that a copy of the draft be forwarded to the White House before it was shared with other senior officials. We were explicit in saying that we were not soliciting proposed edits; we just didn't want the administration surprised when we issued the paper. Despite those caveats, a series of calls from the White House continued to pour in asking us to revise or withdraw the paper. John McLaughlin was on the receiving end of one such call, from a testy Scooter Libby asking for more revisions. The answer was no-we would make no more revisions. Jami Miscik received the brunt of those calls. She, too, stood firm. Jami believed that she had pushed her a.n.a.lysts to ensure they employed every a.n.a.lytic best practice and that no solid reporting had been ignored. But she would not go beyond where the intelligence took us.

After Steve Hadley called Jami from the NSC, wanting to engage her in yet another discussion on the paper, she stormed into my office and said she would resign before she would delay or amend the paper again. Completely supportive of her, I picked up my white secure telephone and punched Hadley's number. "Steve," I said, "knock this off. The paper is done. It is finished. We are not changing it. And Jami is not coming down there to discuss it anymore."

Message received. A day or two later Jami was at the White House-for an entirely different reason-and got word that the president wanted to see her. He clearly had heard about the flap and asked if "his guys" had "stepped over the line." Not wanting to prolong the controversy, Jami told me that she a.s.sured the president that it was nothing that we couldn't handle.

On January 28, 2003, the paper was published. So what did it say? Our a.n.a.lysts believed that there was a solid basis for identifying three areas of concern with regard to Iraq and al-Qa'ida: safe haven, contacts, and training. But they could not translate this data into a relationship where these two ent.i.ties had ever moved beyond seeking ways to take advantage of each other.

The intelligence told us that senior al-Qa'ida leaders and the Iraqis had discussed safe haven in Iraq. Most of the public discussion thus far has focused on Zarqawi's arrival in Baghdad under an a.s.sumed name in May of 2002, allegedly to receive medical treatment. Zarqawi, whom we termed a "senior a.s.sociate and collaborator" of al-Qa'ida at the time, supervised camps in northeastern Iraq run by Ansar al-Islam (AI).

AI, a radical Kurdish Islamic group, was closely allied to al-Qa'ida. Kurdish Islamists and al-Qa'ida had come together in the summer of 2000 to create a safe haven for al-Qa'ida in an area of northeastern Iraq not under Iraqi government control, in the event Afghanistan was lost as a sanctuary. The area subsequently became a hub for al-Qa'ida operations. We believed that up to two hundred al-Qa'ida fighters began to relocate there in camps after the Afghan campaign began in the fall of 2001. The camps enhanced Zarqawi's reach beyond the Middle East. One of the camps run by AI, known as Kurmal, engaged in production and training in the use of low-level poisons such as cyanide. We had intelligence telling us that Zarqawi's men had tested these poisons on animals and, in at least one case, on one of their own a.s.sociates. They laughed about how well it worked. Our efforts to track activities emanating from Kurmal resulted in the arrest of nearly one hundred Zarqawi operatives in Western Europe planning to use poisons in operations. What was even more worrisome was that by the spring and summer of 2002, more than a dozen al-Qa'idaaffiliated extremists converged on Baghdad, with apparently no hara.s.sment on the part of the Iraqi government. They had found a comfortable and secure environment in which they moved people and supplies to support Zarqawi's operations in northeastern Iraq.

More al-Qa'ida operatives would follow, including Thirwat Shihata and Yussef Dardiri, two Egyptians a.s.sessed by a senior al-Qa'ida detainee to be among the Egyptian Islamic Jihad's best operational planners, who arrived by mid-May of 2002. At times we lost track of them, though their a.s.sociates continued to operate in Baghdad as of October 2002. Their activity in sending recruits to train in Zarqawi's camps was compelling enough.

There was also concern that these two might be planning operations outside Iraq. Credible information told us that Shihata was willing to strike U.S., Israeli, and Egyptian targets sometime in the future. Shihata had been linked to terrorist operations in North Africa, and while in Afghanistan he had trained North Africans in the use of truck bombs. Smoke indeed. But how much fire, if any?

Could we prove that this was Iraqi complicity with Zarqawi and the two Egyptian Islamic Jihad operatives? No. Do we know just how aware Iraqi authorities were of these terrorists' presence either in Baghdad or northeastern Iraq? No, but from an intelligence point of view it would have been difficult to conclude that the Iraqi intelligence service was not aware of their activities. Certainly, we believe that at least one senior AI operative maintained some sort of liaison relationship with the Iraqis. But operational direction and control? No.

In the laborious exercise undertaken by a.n.a.lysts to understand the history of a potential Iraqal-Qa'ida relationship, they went back and doc.u.mented the basis of a variety of sources-some good, some secondhand, some hearsay, many from other intelligence services. There were, over a decade, a number of possible high-level contacts between Iraq and al-Qa'ida, through high-level and third-party intermediaries. Our data told us that at various points there were discussions of cooperation, safe haven, training, and reciprocal nonaggression.

During the mid-1990s, Sudanese national Islamic Front Leader Hasan al-Turabi reportedly served as a conduit for Bin Ladin between Iraq and Iran. Turabi in this period was trying to become the centerpiece of the Sunni extremist world. He was hosting conferences and facilitating the travel of North Africans to Hezbollah training camps in the Bekaa Valley, in Lebanon. There was concern that common interests may have existed in this period between Iraq, Bin Ladin, and the Sudanese, particularly with regard to the production of chemical weapons. The reports we evaluated told us of high-level Iraqi intelligence service contacts with Bin Ladin himself, though we never knew the outcome of these contacts.

A senior al-Qa'ida detainee told us in 2002 that he believed it unlikely that Bin Ladin would ally himself with Baghdad and thereby compromise al-Qa'ida's mission and independence. He also said that several of Bin Ladin's lieutenants had urged cooperation with Iraq, believing that the benefit of possible training, safe haven, and help with al-Qa'ida's WMD efforts outweighed any risks to al-Qa'ida's independence. According to the detainee, Saddam became more interested in al-Qa'ida after the East Africa and Cole Cole bombings. But certainly by that time, al-Qa'ida had demonstrated its prowess to conduct conventional attacks, and was well established in its sanctuary in Afghanistan. bombings. But certainly by that time, al-Qa'ida had demonstrated its prowess to conduct conventional attacks, and was well established in its sanctuary in Afghanistan.

The one possible connection that a.n.a.lysts viewed as most disturbing was training. There were solid reports from senior al-Qa'ida members that raised concerns about al-Qa'ida's enduring interest in acquiring chemical and biological expertise from Iraq. In the public debate that has since occurred, this has now all come down to the recantation of an individual named Ibn Sheikh al-Libi. A senior military trainer for al-Qa'ida in Afghanistan, al-Libi was detained in late 2001 and transferred into military custody in Afghanistan in early January of 2002. At the time, he was the highest ranking al-Qa'ida member in U.S. custody.

We believed that al-Libi was withholding critical threat information at the time, so we transferred him to a third country for further debriefing. Allegations were made that we did so knowing that he would be tortured, but this is false. The country in question understood and agreed that they would hold al-Libi for a limited period, and then return him to U.S. military custody, where he would be registered with the International Committee of the Red Cross.

In the course of questioning while he was in U.S. custody in Afghanistan, al-Libi made initial references to possible al-Qa'ida training in Iraq. He offered up information that a militant known as Abu Abdullah had told him that at least three times between 1997 and 2000, the now-deceased al-Qa'ida leader Mohammad Atef had sent Abu Abdullah to Iraq to seek training in poisons and mustard gas. Another senior al-Qa'ida detainee told us that Mohammad Atef was interested in expanding al-Qa'ida's ties to Iraq, which, in our eyes, added credibility to the reporting.

Then, shortly after the Iraq war got under way, al-Libi recanted his story. Now, suddenly, he was saying that there was no such cooperative training. Inside CIA, there was sharp division on his recantation. It led us to recall his reporting, and here is where the mystery begins.

Al-Libi's story will no doubt be that he decided to fabricate in order to get better treatment and avoid harsh punishment. He clearly lied. We just don't know when. Did he lie when he first said that al-Qa'ida members received training in Iraq or did he lie when he said they did not? In my mind, either case might still be true. Perhaps, early on, he was under pressure, a.s.sumed his interrogators already knew the story, and sang away. After time pa.s.sed and it became clear that he would not be harmed, he might have changed his story to cloud the minds of his captors. Al-Qa'ida operatives are trained to do just that. A recantation would restore his stature as someone who had successfully confounded the enemy. The fact is, we don't know which story is true, and since we don't know, we can a.s.sume nothing.

The additional context I had to consider was this: the kind of training al-Qa'ida may have been pursuing with Iraq in the chemical and biological arena was part of a larger, more robust and compartmented WMD program that al-Qa'ida was pursuing and continues to pursue. It is a program sanctioned and directed by the senior leadership. Would they have sought to attain building blocks from more sophisticated programs? My view at the time was that it was completely possible.

Did we look at Zarqawi's operations at the lower-level poisons facility in northeastern Iraq as part of al-Qa'ida's intention to both use these lesser capabilities and also obscure their more important and lethal programs? Of course, you can pull out the al-Libi recantation and say, "You see, this was all hyped." Yet if you ignore the Iraqi context we were operating in with regard to al-Qa'ida's pursuit of WMD capability, you end up missing the larger and more important picture. This was my mind-set. Run it all down, put all the concerns on the table, and give everybody your best judgment.

There was more than enough evidence to give us real concern about Iraq and al-Qa'ida; there was plenty of smoke, maybe even some fire: Ansar al-Islam; Zarqawi; Kurmal; the arrests in Europe; the murder of American USAID officer Lawrence Foley, in Amman, at the hands of Zarqawi's a.s.sociates; and the Egyptian Islamic Jihad operatives in Baghdad. But for some in the administration, it was never enough. They had pushed the data farther than it deserved. They made command linkages where we could not see them. They sought to create a connection between Iraq and the 9/11 attacks that would have made WMD, the United Nations, and the international community absolutely irrelevant. The first problem is that case was never, ever true. The second problem is that in trying to make more out of the case, advocates ended up undermining the case we had. People just stopped listening.

It was during this period that we dealt with another high-profile issue. Reports dating back to late 2001 alleged that one of the 9/11 hijackers, Mohammed Atta, might have met with Ahmad Khalil al-Ani, a member of the Iraqi intelligence service, in Prague just months before the 2001 attack. The White House, Department of Defense, and CIA were all intensely interested in the allegation. If it could be shown that Iraq was an active partic.i.p.ant in the planning for the 9/11 attacks, there would be no question regarding an immediate effort to oust Saddam.

We devoted extraordinary effort to the issue but could never find any convincing evidence that the visit had happened. In fact, over time the intelligence suggesting such a meeting eroded. Proving something didn't didn't happen is problematic, but in this case, we and the FBI concluded that such a meeting was highly unlikely. Nonetheless, we kept being asked to reinvestigate the matter, and while doing so, we kept hearing highly placed officials, including the vice president, say on television that it was "pretty well confirmed" that the visit had occurred. By May of 2002, FBI and CIA a.n.a.lysts voiced increased skepticism that these meetings had taken place. The case for the meetings continued to weaken from that time forward. happen is problematic, but in this case, we and the FBI concluded that such a meeting was highly unlikely. Nonetheless, we kept being asked to reinvestigate the matter, and while doing so, we kept hearing highly placed officials, including the vice president, say on television that it was "pretty well confirmed" that the visit had occurred. By May of 2002, FBI and CIA a.n.a.lysts voiced increased skepticism that these meetings had taken place. The case for the meetings continued to weaken from that time forward.

It is my understanding that, in 2006, new intelligence was obtained that proved beyond any doubt that the man seen meeting with the member of the Iraqi intelligence service in Prague in 2001 was not not Mohammed Atta. Mohammed Atta.

A second possible linkage to 9/11 and Iraq involved an Iraqi national named Shakir who worked at the airport in Kuala Lumpur as a part-time facilitator for Arab visitors, a job he had obtained through an Iraqi emba.s.sy employee. In January 2000, Shakir facilitated the travel of 9/11 hijacker Khalid al-Mihdhar from the airport. Shakir's immediate departure from Malaysia one week after helping al-Mihdhar through the airport, and previous travel and contacts with extremists, raised red flags. After months of exhaustive a.n.a.lytic work, we could not establish that Shakir was an Iraqi agent.