Friday, June 11, 2004
CIA misses deadline to fix analysis flaws
Source: Douglas Jehl, New York Times - 06/10/04
The Central Intelligence Agency has yet to come up with a plan to address what senior officials have described as a major flaw in its operations, despite a pledge four months ago that the problem would be resolved within 30 days.
The problem, which contributed to errors in the agency’s prewar estimates on Iraq, is rooted in procedures that severely limit how much information about human sources is shared with analysts who produce intelligence assessments, according to senior intelligence officials.
In a Feb. 11 speech, a senior CIA official, Jami Miscik, described the problem as an example of ‘‘imperfections in our system’’ and said that George Tenet, the director of central intelligence, had given subordinates 30 days ‘‘to devise a permanent and lasting solution.’’ But this week, a senior intelligence official said that a team headed by the agency’s executive director, A. B. Krongard, was ‘‘still working out the modalities’’ of a new arrangement.
A senior intelligence official said this week that the recently announced departures of Tenet and James Pavitt, the deputy director for operations, meant that it was unlikely that the new arrangements would be worked out soon.
The difficulty of working out a solution reflects a deep gulf between the CIA’s operations directorate, which recruits and run spies around the world, and its intelligence directorate, which is in charge of sifting intelligence from those spies and from satellites and eavesdropping devices and drawing broad conclusions.
In the case of Iraq, analysts who wrote reports stating that the country possessed illicit weapons did so without knowing that some of the central sources they cited were defectors linked to Ahmad Chalabi’s Iraqi National Congress, according to senior intelligence officials.
[..]
In these and other cases, the prewar assessments about Iraq’s illicit weapons were based on reports from sources who did not have firsthand information about what they described.
That fact, too, was sometimes known by intelligence officers but was rarely shared with intelligence analysts, ac cording to the senior intelligence officials.
In her speech, Miscik said that ‘‘the biggest lesson’’ to have emerged from the apparent mistakes in the agency’s prewar assessments on Iraq was ‘‘the importance of getting the analyst as much information as possible about a source’s access.
‘‘Analysts can no longer be put in a position of making a judgment on a critical issue without a full and comprehensive understanding of the source’s access to the information on which they are reporting,’’ Miscik said in the speech to intelligence analysts at the agency, a copy of which was provided to The New York Times.
[..]
Under the longstanding arrangements that Miscik said would be changed, the directorate of operations typically shields information about sources’ identity, motivations and even access to information from the directorate of intelligence.
[..]
In her speech, Miscik did not cite specific examples in which inadequate sharing of information had led to mistakes in the prewar intelligence on Iraq.
But she said an internal review of that intelligence had found cases in which a single source had been described in different ways, so that an intelligence analyst might believe the same information actually had come from multiple sources.
‘‘In an age where policymakers are relying on intelligence to inform their decision-making, we cannot let these imperfections in our system continue,’’ Miscik said in the speech.
(end of excerpt)