The CIA’s own evaluation admits it couldn’t rule out precognition

This page focuses on one specific section of a document already covered in full on this site: An Evaluation of Remote Viewing: Research and Operational Applications, prepared by the American Institutes for Research at the CIA’s request in September 1995. The full report and its headline findings, the Utts and Hyman conclusions and the program’s termination, are covered on this site’s main STARGATE page. This page isolates a narrower, methodological passage that rarely gets quoted.

Remote viewing sessions at SRI and later SAIC typically followed a free-response format: a viewer describes a target with all known sensory channels blocked, then receives feedback confirming or denying accuracy. The evaluation’s own text explains why this design creates an unavoidable ambiguity: in most cases the viewer eventually receives feedback in which he or she learns the correct answer, thus making it difficult to rule out precognition as the explanation for positive results, whether or not there was a sender actively looking at the target during the session.

The report distinguishes this from telepathy and clairvoyance specifically because precognition is, in its own words, easier to identify: it is possible to identify apparent precognition by asking someone to describe something for which the correct answer isn’t known until later in time. The evaluators are not asserting precognition happened. They are stating, in a formal CIA-commissioned scientific evaluation, that their own experimental design could not cleanly rule it out as an explanation for the statistically significant results the program produced.

This establishes that the CIA’s own independently commissioned scientific evaluation of a twenty-three-year government program formally acknowledged, in writing, that a portion of its documented statistical results could not be attributed with confidence to remote viewing of a present-tense target rather than precognition of information the viewer would only later be told. This is the evaluation’s own stated methodological limitation, not an outside interpretation of it.

What this does not establish is that precognition occurred, or that any specific session demonstrates it. The report treats this as an unresolved confound in its experimental design, not a confirmed finding. The honest reading is narrower and more interesting than either extreme: a formal government evaluation could not, using its own methods, distinguish an ordinary explanation from a genuinely strange one.