Before the 1991 Gulf War, they said that Saddam didn't have any WMD programs beyond the chemical weapons he used on the Iranians and the Kurds. He did.

After the Gulf War, they consistently said Saddam had active WMD programs. They said it all through the Clinton Administration, which is why Clinton (to his credit) isn't tearing into Bush for exaggerating the intel.

The CIA and the IC (Intel Community) failed to predict 9-11. They consistently overestimated the threat of the former Soviet Union for decades. Now, they're concerned because someone is going to hold an investigation at intelligence failures, so they're just trying to cover their collective asses.

From the NY Times: The CIA: Method and Madness

For a more balanced perspective on the IC and how it actually function, there's a good letter on National Review Online:

“Secondly, the intelligence community must naturally err on the side of pessimism and alarmism. The cardinal rule in military intelligence is to estimate the worst your enemy could possibly do, not what he probably is willing to do, or could economically do, or is likely to do. Worst-case analysis is the rule of the day. How could it be otherwise?

“I mean, here people are simultaneously saying 9/11 should have been anticipated, which would have required making seriously worst-case analyses about the threat posed by various low-level riff-raff, while at the same time making worst-case (instead of probable-case) analyses of the threat of Saddam Hussein was 'irresponsible.' Well, which is it? Which rule should the spooks follow?

“Additionally, the CIA was trained during the Cold War, when almost comically worst-case analyses of the Soviet threat were entirely routine. In the 80s we were told the Soviet threat necessitated enormous efforts -- MX missiles, 600-ship Navy, Los Angeles class U-boats, Star Wars, et cetera -- while in truth the Soviet state was at that time starting a spectacular economic and military power dive out of which it has still not pulled.

“Let us not even get into the late '50s missile gap, and the frantic rush to build the H-bomb, or even Einstein's letter to FDR warning of the threat of a German A-bomb, which speak for themselves in terms of spectacular mis-estimation of enemy capabilities.

“My point is not to fling contempt on the intelligence community, but to point out that when it functions as we expect -- nay demand -- it will routinely provide threat estimates which are exaggerated, and regularly provide threat estimates which do so wildly.

“To make exact threat estimates requires either a time machine, so you can take advantage of the hindsight the President's critics are all indulging in so shamelessly, or else unbelievably precise knowledge of those very activities which our enemies are willing to invest the most in hiding.

You can "coulda, woulda, shoulda" all you want, but the reality of the matter is that it has always been difficult to create an accurate assessment of what happens inside of secretive totalitarian countries. In the end, the "consensus" is often little more than an educated guess based on fragmented evidence.