Pentagon Allocates $30M for AI‑Enhanced Polygraph Program

The Pentagon has earmarked $30.3 million for a new AI‑powered lie detector program, called Polygraph+, that the Defense Counterintelligence and Security Agency (DCSA) plans to use for employee vetting and insider‑threat detection. The initiative will focus on scoring algorithms that use AI and machine learning and on a technique called standoff sensing, which takes physiological readings without attaching a device.

Key points

• DoD requests $30.3 million over five years for Polygraph+.
• Program centers on AI‑driven scoring and standoff sensing.
• DCSA will deploy it for vetting employees and insider‑threat detection.
• Polygraph technology has been debunked; accuracy claims range 80‑94 % but real‑world error remains high.
• Minority groups and trained countermeasures can skew results.

What happened

The budget document, first reported by Inside Defense, details a five‑year plan to modernize federal polygraph and credibility assessment technologies. Polygraph+, also called Polygraph Next, will be run by the DCSA, which conducts background checks for the federal government. The new technology will be used for vetting prospective employees and “insider threat detection.” The request is still pending congressional approval.

The DoD’s interest in polygraphs has surged under Secretary Pete Hegseth, who has increased reliance on the tests to locate alleged leaks. In September, The New York Times reported that about 50 officers on the Joint Staff had been given polygraph tests after coverage of depleted U.S. weapons stockpiles in the war with Iran. The Pentagon’s push reflects a broader effort to control information flow within the agency.

Polygraph technology has barely changed since its invention in the 1920s. Examiners use blood pressure, pulse, breathing, and sweat measurements to gauge truthfulness by comparing baseline questions to target questions. The federal government conducts tens of thousands of tests a year, yet the polygraph has been repeatedly debunked. The 1983 Office of Technology Assessment concluded that there was very limited evidence supporting the polygraph for employee screening, and the 2003 National Research Council said evidence on its efficacy was “weak at best.”

Human observers spot lies about half the time without assistance. The American Polygraph Association claims the device is 80‑94 % accurate, but the NRC report warned that even that level of accuracy could produce many false positives when applied to 2.8 million DoD staff.

The backstory

The Pentagon’s current emphasis on polygraphs coincides with a series of high‑profile leaks and internal investigations. In 2023, the Defense Innovation Unit (DIU) launched an open call for deception‑detection products. Two companies were selected to build prototypes: Presage Technologies, which claims to measure heart rate and breathing with standard cameras, and Altec Research, a medical sensor firm exploring non‑contact sensing. A DIU screenshot of Altec’s prototype shows it tracks head movement, facial skin temperature, and pore activity.

Despite modern AI advances, the polygraph’s core methodology remains unchanged. The DCSA’s upcoming program will therefore rely on new scoring algorithms and standoff sensing to try to overcome longstanding limitations.

Why the DoD is investing

The Pentagon’s budget request signals a belief that technology can improve a tool that has historically failed to meet scientific standards. By applying machine‑learning scoring, the agency hopes to reduce subjectivity and increase consistency across examiners. The focus on standoff sensing also addresses the logistical burden of attaching sensors to every subject, potentially expanding the test’s reach.

However, the cost–benefit calculus is uncertain. The $30.3 million investment may not translate into a meaningful reduction in false positives or improved detection of insider threats. The DoD’s reliance on a contested technology could also erode trust among potential recruits and contractors.

Risks and trade‑offs

The program faces significant ethical and operational challenges. The polygraph’s documented bias against minority groups and the ease with which trained countermeasures can be employed raise concerns about fairness and legal defensibility. Even if AI scoring improves accuracy, the sheer scale of DoD personnel—2.8 million employees—means small error rates could still flag thousands of individuals.

the lack of congressional approval and the absence of a clear implementation plan mean the program could stall or be scaled back. If the technology fails to meet expectations, the DoD may need to revert to traditional screening methods or seek alternative deception‑detection solutions.

What builders and businesses should do

• Track congressional hearings on the Polygraph+ budget request.
• Evaluate the technical claims of AI scoring and standoff sensing against independent studies.
• Prepare for potential integration of Polygraph+ data into contractor vetting workflows.
• Review compliance requirements for handling biometric data in defense contracts.
• Monitor developments from DIU prototypes to identify emerging commercial alternatives.

The bottom line

Watch for the next congressional budget cycle to see if the $30.3 million request is approved. The outcome will shape how the DoD screens personnel and could set a precedent for AI‑driven deception detection in other federal agencies.