Back to Index

US: The intelligence process

Miles Seeley, who spent his career in the CIA, writes: "As the Senate prepares to investigate the way various government agencies processed intelligence data about Iraq prior to the invasion, it might be useful to consider exactly how the system works. Following is how it worked in my day.

In the CIA, basic intelligence is collected from satellite or electronic sources, or by and from human sources. Photo interpretation and NSA translation of intercepts is the next step for data from non human sources. Finally, selected portions are gathered together for transmittal to the end users- the upper echelons of the White House, the Departments of Defense and State, and others. At all stages of this process, some material is judged relevant, some of lesser importance. The judgements are made by skilled professionals operating under guidelines from above; i.e. they select data about WMD but not the condition of crops.

For so-called HUMINT (human intelligence) the system works in similar but not identical ways. The field case officer collects information from his sources, and this may be in the form of documents or photos or simply verbal reports. The case officer makes a preliminary judgement of the relevance and truthfulness of the material, based on his knowledge of his source and the targets. The report is then sent to CIA Headquarters, where it first arrives at the country desk, usually staffed by officers who have served in the area. They in turn send the reports on to the analysts, who are completely separate and work independently. The analysts screen the material and judge it against all other material on the subject. They look at both overt and covert reports, and previous data collected from all sources. They then grade the material for accuracy and relevance, and their judgement may be quite different from that of the field or desk case officers. The material is then screened again and assigned a priority, the highest being information that should be included in the daily briefing of the President ( and, usually, the Secretaries of Defense and State and others).

Now comes the trickiest part, the final judgement of the most senior officials about what information suits their immediate needs. This is where politics comes in, and the particular agendas of these officials. Thus, if the administration is ramping up for an invasion and wants justification for it, intelligence data useful for that would be highlighted, and data that does not support that agenda is ignored or given lower priority. The field case officer, the desk officer, and the analyst have no agenda, in theory at least.

There is an enormous amount of raw information coming in to various agencies and departments every day, much of it seemingly contradictory. For example, the CIA might conclude it has no solid, verifiable information about the existence of WMD in Iraq, although it has many hints of it. The Pentagon might not be satisfied, since its agenda is to confirm the presence of MD as justification for invasion. So some people in the Pentagon might set up a small group to look again at all the raw data, concentrating on that which "proves" the presence of WMD and ignoring that which does not. This is not an intelligence process, it is a political one. Since there is so much information, for example from Iraqi exiles with little access to relevant data but a strong agenda, the small group in the Pentagon can probably find plenty of material to support their perceived needs.

If the final conclusions are shown to be wrong in the end, whose fault is it? Usually, the press (and the high officials who skewed the data) blame the CIA. That is easy to do and enables the people really at fault to do what they always do- CYA, in the vernacular, or Cover Your Ass.

Ronald Hilton - 6/7/03