At the Center of the Storm: My Years at the CIA

by George Tenet;Bill Harlow

  What they didn’t know at the time, of course, was that Pincus had learned about the Niger mission from Ambassador Joseph Wilson, the man CPD had asked to undertake the trip.

  How did the trip happen? Several of our briefers had received questions not only from the vice president but also from the State Department and DOD about a February 2002 Defense Intelligence Agency report that first raised the possibility of Iraq having sought uranium from Niger. “What more do you know about this?” they were asked. “Hardly anything,” was the answer. Midlevel officials in CPD decided on their own initiative to see if they could learn more. Someone had the idea that Joe Wilson might be a good candidate to look into the matter. He’d helped them on a project once before, and he’d be easy to contact because his wife worked in CPD. Wilson agreed and undertook the assignment without compensation. Only his expenses were reimbursed.

  Critics have subsequently suggested that our selecting Wilson demonstrated that the Agency had it in for the administration. After all, wasn’t he a supporter of the Democrats? I would argue that his selection illustrates that Agency officers often don’t give a second thought to U.S. domestic politics. The report that Saddam might be getting yellowcake from Niger was not an issue of left or right—it was either right or wrong.

  Not surprisingly, local officials in Niger denied illegally selling uranium to Iraq. Wilson didn’t even write up a report; he gave an oral briefing to two CIA analysts at his home one evening over Chinese takeout food. Their summary of his remarks said that the officials denied selling yellowcake to Iraq but that one official admitted Iraq had been seeking expanded trade relations with Niger. The presumption was that the only thing Niger had worth trading was yellowcake.

  This unremarkable report was disseminated, but because it produced no solid answers, there wasn’t any urgency to brief its results to senior officials such as the vice president. Had the vice president been in Washington at the time, his personal PDB briefer might have mentioned it, but as it happened, Cheney was on a ten-day overseas trip when the report came out. By the time he returned to Washington, there were undoubtedly more pressing things to brief him on. As far as we could tell, the Wilson summary was never delivered to Cheney. In fact, I have no recollection myself of hearing about Wilson’s trip at the time.

  Pincus’s story, which ran in the Washington Post on June 12, revived interest in the State of the Union address and yellowcake, and for several days thereafter the rest of the news media chased the issue, trying to sort out who had said what to whom—and how those sixteen words had gotten into the speech. Several follow-on stories by Pincus cited sources close to the vice president complaining that CIA had “failed” to keep them informed. It was pretty clear that some anonymous staffers in the vice president’s office were trying to make sure that if there were any fallout over the issue, CIA would solely be held at fault. This became a familiar theme for us.

  Then the issue seemed to have died again. And for me, the matter was nowhere near the top of my list of things to worry about that spring. Yes, the temperature was rising, but at any given time dozens of such issues are bubbling in Washington. Try as you might, you never know which mini-crises will subside and which will boil over. When I called Condi that June to express my concern over the matter, I was troubled by the weak intelligence that underlay the phrase, not with Joe Wilson. I have to admit, I did not see trouble looming when I first learned that Wilson’s wife, Valerie, was a CIA employee. I did not view that as a big deal or a political vulnerability, or much of anything, for that matter. Condi called several days after my call to say the White House would not be issuing any statements saying that the Niger material should not have been used. Condi made it clear to me that this was not her decision.

  Sunday morning, July 6, dawned a typical Washington summer’s day. I tried not to go into work on Sundays so I could spend as much time as possible with my family. But work always came to me. My ever-present security detail delivered a stack of overnight cable traffic, intelligence analysis on critical issues, and a thick package of clippings from the morning’s newspapers, called the Media Highlights, with stories relating to intelligence. Prominently displayed in the Media Highlights was a column written by Ambassador Wilson that appeared in the morning’s New York Times. Apparently, he had decided that feeding anonymous stories to Kristof and Pincus had not achieved his goals, so this time he outed himself in an Op-Ed piece titled “What I Didn’t Find in Africa.”

  While the earlier Kristof and Pincus articles had touched off brush fires, the Wilson Op-Ed and subsequent TV appearances ignited a firestorm. I’d been around Washington long enough to know that when you attach a name to an allegation, the story has much greater traction. If there were any doubt, it was removed when I tuned in to NBC’s Meet the Press that morning and saw the guest host, Andrea Mitchell, interview Joe Wilson on his allegations that the administration had ignored his findings and hyped the Niger information even though, in his estimation, they “knew” the claim not to be true.

  By Monday morning virtually every major news organization was chasing the story. Ari Fleisher, the soon-to-depart White House spokesman, was swamped with questions at his early morning press “gaggle,” an on-the-record but off-camera media briefing. Ari told reporters that there was “zero, nada, nothing” new in the weekend’s coverage other than the fact that Wilson’s name was now attached to the allegations. He was pressed on whether the White House still stood by the words in the “SOTU”—Washington-speak for the State of the Union address.

  Fleisher danced around that, but later that day—after the president, the White House staff, and the traveling press corps departed on a trip to Africa—Ari’s staff finally released a brief statement acknowledging that the uranium language should not have been included in the speech. The White House had finally gotten around to saying the obvious—saying, indeed, what I had said to Condi Rice a few weeks before. I know of no meeting that was convened to come to this decision. The White House staff simply read the tea leaves after Joe Wilson’s weekend media appearances and decided to commit truth.

  That should have ended the matter. The White House essentially admitted that “mistakes were made,” “we’re sorry,” and “let’s move on.” Each day brought fresh stories quoting anonymous officials pointing fingers at each other’s organizations. Pundits began opining that the White House had deliberately misled the American people. The word “lied” was bandied about by administration critics.

  Presidential overseas trips are especially likely times for self-inflicted crises to spring up. A huge press contingent and many staffers travel with the president—so many people that two 747s, Air Force One and its twin, are not big enough to handle them all. The White House staff spends too much time cooped up together, and they tend to get one another spun up with the latest tales of what they are hearing from back home. Meanwhile, the press contingent is hungry for any tidbits or insider bickering to report. In the hothouse environment of Air Force One an “us against them” attitude often leads to badly thought through reactions.

  As it turned out, I had some traveling to do of my own: a long-standing speaking engagement in Sun Valley, Idaho. This was at an event sponsored by Herbert Allen, whose investment banking company specializes in working with major figures in the entertainment, communications, and technology fields.

  I made an hour-long off-the-cuff presentation in the Sun Valley Lodge’s conference room about the state of the world as I saw it. It was my second appearance in front of this crowd; I enjoyed the informal banter in the subsequent question-and-answer period with the eclectic group of participants.

  At one point, one of the attendees, NBC anchorman Tom Brokaw, suggested that I should be making this same kind of presentation on national TV. In front of the assembled group, which included some of his competitors, he offered me an opportunity to do so on NBC.

  “Yeah, Tom,” I said, “it’s always been my dream to be grilled by you on national TV.”


  “Well, George,” he replied, “you know we are in Sun Valley, and they call this ‘the place where dreams come true.’”

  Tom earned a big laugh—but no interview.

  In addition to offering a chance to speak to an influential group of people, the trip also provided me an opportunity to take a day or two off in a beautiful setting. After checking with our ethics attorneys and agreeing to pay her expenses, I was able to bring Stephanie along for what I hoped would be a relaxing couple of days. But there was to be no relaxation. Almost from the moment I arrived in Idaho, I’d been bombarded with calls from Washington about growing concerns regarding the State of the Union controversy. Now, rather than enjoying the mountain trails and streams, I found myself fielding a never-ending flow of phone calls from headquarters telling me of the latest sniping going on across the Potomac and now across the Atlantic.

  Stephanie and I were staying in a room in the main lodge that was said to have once been occupied by Ernest Hemingway. Unlike “Papa” Hemingway, though, we had to take over the adjoining room as well for a “command post.” That was standard procedure. Whenever I traveled, even to a garden spot like Sun Valley, a team of communicators would arrive ahead of me and set up an office with sophisticated satellite communications equipment, allowing me to be in touch with national command authorities and to receive highly classified voice and data transmissions. The team would work in shifts to ensure that someone was always in touch with our headquarters back home. When taking trips with multiple stops, communications teams would have to leapfrog ahead of me, moving hundreds of pounds of equipment that would permit encrypted communications as soon as I stepped off the airplane at the next destination.

  The communications this time were virtually nonstop. The classified fax machine kept humming, spitting out news stories, briefing transcripts, and editorials—a barrage that made it crystal clear this story wasn’t going away soon.

  Finally, at one point I decided that I had had enough. I called Steve Hadley at the White House. “We need to put an end to this,” I told him. As I had explained to Condi in my call to her a few weeks before, including the uranium language in the State of the Union speech had been a mistake. Now, I said, I had decided that I would issue a statement accepting responsibility for the Agency’s shortcomings in allowing the uranium language to make it into the speech. I would stand up and take the hit. Obviously, the process for vetting the speech at the Agency had broken down. We had warned the White House about the lack of reliability of the assertion when we had gotten them to remove similar language from the president’s October Cincinnati speech, and we should have gotten that language out of the SOTU as well. It was because of my failure to fully study the speech myself that I took responsibility. We owed it to the commander in chief, and we had failed him, and now, I told Hadley, was the time to own up to that.

  Hadley candidly responded that the process had not worked well at the White House, either—and that they would stand up with us. “It will be shared responsibility, George,” he told me. For that reason, I fully expected Condi Rice to publicly state that she joined me in accepting responsibility.

  I wasn’t just being magnanimous. Part of the fault truly was mine. The day before the State of the Union, I was at a Principals’ meeting in the White House Situation Room, a place where it seemed I spent more time than in my own home in recent years. As the meeting broke up, several of us were handed copies of a draft of the forthcoming speech. I remember going back to headquarters and giving the draft to one of my special assistants, unread, and asking that it be put “into the system for review.”

  I gave it no further thought. As always, other crises were banging on the door, but I fully expected that if there were any problems with the State of the Union draft, someone would have come and alerted me. That’s exactly what had happened with the Cincinnati speech the previous fall. On another occasion, involving the 2002 State of the Union speech, my chief of staff, John Moseman, and spokesman, Bill Harlow, intervened at the last moment to stop the president’s speechwriters from including language about the number of terrorists believed to have been trained in Bin Ladin’s camps in Afghanistan, a number that was tens of thousands beyond what we thought true. Moseman called the NSC staff and said, “Look, if the president goes out and says that and tomorrow media call us and ask if we agree with the number, Harlow is going to have to say no. The number was corrected at the last minute—so late that an advance text copy of the speech put out at a background briefing at the White House that evening still contained the unsupportable tally.

  In early 2003, though, the same system and same people that had rescued the president from incorrect assertions in previous speeches failed to catch the troublesome language in the State of the Union. Later, in trying to find out why alarm bells hadn’t gone off, I was told that Alan Foley, head of WINPAC, had focused on clearing the speech for “sources and methods,” rather than for substance. In other words, as long as the language didn’t give away any secrets about how the intelligence was collected, they didn’t worry about whether we believed the assertions in the speech were accurate. That was a terrible mistake. Our job was never to clear solely for sources and methods, but also for substance. And the last time I looked, as good as the British intelligence service is—and it is very good—it does not work for the president of the United States.

  On the morning after I talked with Steve Hadley, I called Washington, pulled Bill Harlow out of the morning staff meeting, and told him that I had decided to issue a statement taking our share of the blame for the mix-up. I gave him a sense of how I wanted the statement to go, and read him a few opening paragraphs I had scribbled on a yellow legal pad overnight, since I hadn’t been able to sleep.

  My instructions were clear: “I want this statement scrubbed carefully. It has to be as accurate as we can make it. Factual, clear, and no whining allowed.” But more than just saying “we screwed up and we’re sorry,” I wanted to lay out to the extent possible what had happened. The statement also needed to be a roadmap and to convey the clear impression that we never believed the Niger story. Most important, I wanted to say that we regretted having let the president down and that I took personal responsibility.

  My deputy, John McLaughlin, and Bill Harlow labored long and hard trying to construct a statement that would accomplish what we wanted and stand up to scrutiny. It was a painful process. They wrote version after version trying to get the language right, faxing drafts back and forth to me in Idaho and checking with all the appropriate players at CIA headquarters. Among the people they needed to consult was Alan Foley, a senior CIA official who had discussed and eventually cleared the language for the State of the Union with Bob Joseph, a senior NSC official. John and Bill wanted to ensure that they understood Foley’s actions and position, but as it turned out, he was on an official trip to Australia. So there I was in Idaho, coordinating a statement with my staff in Washington while they were reaching out to a key player in Australia, and we were all looking for more incoming flak from the traveling White House in Africa.

  Early in the process, I decided that I wanted to inject some perspective. Yes, it was a bad thing that some of the language drafted for the president’s remarks didn’t rise to the level of certainty that one would expect, but after all, we were talking about a tiny fraction of his speech. That’s when Bill counted and found that we were talking about only “sixteen words”—a phrase that would take on a life of its own. Later some would allege that this handful of words was critical to the decision that led the nation to war. Contemporaneous evidence doesn’t support that, but just try convincing people of that today.

  A better case could be made that the “sixteen words” started an unintended war between the White House and CIA. That was certainly not our intention. If there was such a war, it was largely one-sided. Neither I nor my senior leadership ever considered ourselves at war with the vice president or anybody else.

  At one point, Steve Hadley asked me to cal
l Scooter Libby, the vice president’s chief of staff, to discuss my forthcoming statement. I refused to do so. The statement was to be mine and no one else’s. I’ve subsequently seen reports that Libby and Karl Rove debated what they would like to see in my statement. Perhaps so, but I was unaware of their views at the time.

  Sometime between drafts one and seventeen of my “mea somewhat culpa,” Bill Harlow was interrupted by a call from syndicated columnist Bob Novak. Novak said that two administration sources had told him that the real story on the Joe Wilson trip was that Wilson’s wife worked for the Agency and was responsible for sending her husband. Bill struggled to convince Novak that he had been misinformed—and that it would be unwise to report Mrs. Wilson’s name. He couldn’t tell Novak that Valerie Wilson was undercover. Saying so over an open phone line itself would have been a security breach. Bill danced around the subject and asked Novak not to include her in the story. Several years and many court dates later, we know that the message apparently didn’t get through, but Novak never told Bill that he was going to ignore his advice to leave Valerie’s name out of his article.

  I was amused to hear Novak subsequently say that he is confident that I must have been aware of his call at the time and that if I had only phoned him to tell him not to run the item, he would have complied. I was not aware of Novak’s call. I was consumed with the “sixteen words” flap, wondering if in the next few days I would need to resign or would perhaps be fired. About two weeks after Novak’s column appeared, CIA lawyers sent to the Justice Department a formal notification that classified information may have been inappropriately leaked to the media. CIA lawyers had to make that kind of notification about once a week on average. I was informed after the fact that a “crimes report” had been submitted. I supported the action but had nothing to do with the decision. It’s been suggested I ordered the action to get back at the White House for some reason. This is absurd. At the time we had no idea where the leak had come from but were obligated by law to report it to the proper authorities. I was angered that someone, whether intentionally or not, blew the cover of one or our officers and that they appeared to be implying that some desk-bound analyst at Langley was sending her husband on a boondoggle. This was never the case. Nor can we have outsiders determine who is legitimately undercover, ever—because it suits the politics of the moment. To do so is irresponsible and dangerous.