Pentagon seeks $30.3 million for proposed AI lie detection programme
Polygraph+ would combine machine-learning scoring with remote physiological sensing for employee vetting and insider threat detection. Congress has not approved the funding request.
The US Department of Defense has requested $30.3 million over five years to develop Polygraph+, a proposed programme using artificial intelligence and machine learning in credibility assessments. The request, reported by MIT Technology Review, is awaiting congressional approval.
The Defense Counterintelligence and Security Agency would use the technology to vet prospective employees and detect insider threats. The proposal includes scoring algorithms and “standoff sensing”, which would take physiological readings without attaching sensors to a person. The specific technologies have not been identified.
The proposal comes amid increased use of polygraph tests in Pentagon leak investigations. The New York Times reported that around 50 Joint Staff officers were tested after coverage of depleted US weapons stockpiles. The report does not establish that those tests prompted Polygraph+.
Polygraphs assess changes in measures such as blood pressure, pulse, breathing and sweating. Researchers cited by MIT Technology Review question whether these signals can reliably reveal deception. Past assessments have found limited or weak evidence for polygraph screening, and other technology-based detection efforts have not produced reliable results outside laboratory settings.
The programme’s effectiveness and how it would be evaluated or deployed remain unclear. The Defense Counterintelligence and Security Agency did not respond to MIT Technology Review’s request for more information.
