Intelligence Analysis Is Not Scientific Investigation

Several developments this week in stories I’ve been following. I finally have some time to post about them.

Questions about how SARS-CoV-2 jumped from animals to humans and the so-called Havana Syndrome have both been subjects of recent intelligence assessments. Intelligence assessments are different beasts from scientific conclusions. I wrote about that for Scientific American.

Intelligence analysis privileges the credibility of sources; science privileges the analysis of data. Intelligence keeps much of its analysis secret; science publishes and argues in public.

Intelligence analysis is a poor way to investigate the origin of SARS-CoV-2 in humans. Several good scientific studies, linked in my op-ed, point strongly toward the animal market in Wuhan as being the source. Those advocating an origin in a laboratory have put forth a number of scenarios, and very little data. The overall intelligence assessment ranges from “not enough data,” through most agencies’ “low confidence,” to the FBI’s slightly more confidence in a lab leak.

The FBI botched a somewhat similar investigation into the anthrax letters sent after 9/11. There is no reason to believe their assessment, offered without explanation.

The Biden administration used intelligence powerfully in the runup to Russia’s invasion of Ukraine. The intelligence agencies would do well to learn from that experience.

Update: I’m noticing some evidence-free assertions in the comments. Usually I don’t comment on that sort of thing, but because part of the purpose of this post and my op-ed is to clarify what kind of evidence supports what kind of argument, I’ll suggest 1) that if you want to opine, please read the scientific literature; don’t rely on your half-recalled factoids from social media and 2) that you’re contributing to the overflow of misinformation.

Cross-posted to Nuclear Diner