The failure to find weapons of mass destruction in Iraq raises serious questions about America's intelligence community, and the government's use of intelligence. These weapons were our primary rationale for going to war. Americans want to know: how could we have been so wrong?

Intelligence is a murky business. Policymakers want conclusive information, but it is rarely that simple. I have read thousands of intelligence assessments over more than three decades, and they are almost always carefully qualified with phrases like "it is possible that…" or "it is reasonable to assume that...", or words like "may" or "probably." Intelligence analysts are cautious people who deal with competing information and ambiguous sources. Indeed, George Tenet is correct when he says the CIA is faced with, "the unclear, the unknown, and the deliberately hidden." In such a shadowy world there are almost no certainties.

Nonetheless, in the months before war in Iraq, policymakers from President Bush on down expressed certainty that Saddam Hussein had weapons of mass destruction. Were these faulty claims a result of defective work by the intelligence community, or exaggeration by politicians? The answer is likely a little of both.

On the first question, it seems that the intelligence community relied too heavily on uncertain sources: Iraqi defectors and exiles who had an interest in removing Saddam Hussein from power, or inconclusive satellite images and tape recordings like those played by Secretary Powell at the United Nations. We also followed a line of thinking based on old information: Saddam Hussein used chemical weapons in the 1980s and had weapons of mass destruction in the early 1990s, he didn't adequately prove that they were destroyed, and in the post-9/11 world we assumed that he still had the weapons.

Apparently, we lacked direct evidence or good human sources on the ground in Iraq to confirm our suspicions. Instead, we were left with speculation, anecdotes, second or third-hand accounts, photographs of unmarked trucks and buildings, and outdated information. This information was not rigorously evaluated or balanced by competing assessments. It was also not presented to the American people with the proper qualification.

This is where the focus must turn to the policymakers. In my experience, Presidents tend to get the intelligence they want; after all, they appoint leading officials and set priorities for the government. That does not mean the intelligence is wrong; it does mean that it must be viewed with caution. It is human nature for people to look for information that supports their views, and for subordinates to look for information that satisfies their superiors. To some extent, all seven presidents I have worked with looked for – and received – intelligence that justified their preferred policy, and faced the temptation to manipulate that intelligence to sharpen their case.

In the run-up to war, it was no secret that President Bush and his senior advisors suspected the worst of Saddam Hussein, and they were presented with the intelligence assessments that supported their position. Dire conclusions about Iraq's weapons programs rose through the ranks of the government; competing analyses didn't. Uncertain assessments that confirmed the Bush Administration's view of Iraq were taken as incontrovertible evidence. Over a few months, the caveats and qualifiers that normally accompany intelligence assessments were left off of public pronouncements.

No doubt this dramatic breakdown in our collection, analysis and use of intelligence will resonate through the intelligence community and the government for years. It should. The President, his most senior advisors, and the Congress made an assessment of the threat posed by Saddam Hussein that appears to have been incorrect. But questions should reverberate beyond the halls of Washington.

The American people need to think and learn more about intelligence. We need to recognize that all the satellites, agents and analysts in the world cannot often convert our intelligence assessments into "smoking guns" and certitudes: intelligence is less than perfect. We need to consider how that affects the way our Nation makes decisions about war and peace, particularly preemptive actions. We need to consider how we might better organize our government to improve our intelligence. And we need to ensure that intelligence is used as a tool to make good policy – not as a tool to make a policy look good.

In the "the unclear, the unknown, and the deliberately hidden" world of intelligence, caution is critical. Just as we need to find ways to uncover good information to protect ourselves, we need to reject bad information. Most of all, we need to see intelligence for what it is: a tool that can help us take necessary actions, and not a cure-all to keep us from all harm.