The Pentagon wants $30 million to build an AI-powered lie detector
Pentagon plans $30M to build AI‑powered lie detector, hoping tech can spot deception better than old polygraphs.
Why Now
The U.S. Department of Defense requested $30.3 million over five years for a project called Polygraph+ to modernize lie‑detection tech amid concerns over leaks and insider threats.
What Happened
Polygraph+ will use AI/ML scoring algorithms and standoff sensing to read physiological signals without contact. The program is led by the Defense Counterintelligence and Security Agency and is still unapproved by Congress. Existing polygraph accuracy is weak, with studies showing only about 50‑60% human detection rates and significant false‑positive risks.
Why It Matters
If successful, the system could improve vetting of federal employees and deter leaks, but unreliable results could lead to wrongful accusations at scale. The effort reflects a broader push to use AI for security screening, raising questions about privacy and fairness.
The Limitation
The technology’s effectiveness is unproven; polygraphs have historically shown limited accuracy, and AI models may inherit the same uncertainty without a reliable ground truth.
What You Can Do
Follow the Pentagon’s budget updates and review the 2003 NRC report on polygraph efficacy to gauge the project’s feasibility.
Source
Read original sourceWhy we picked this
The Pentagon’s AI-powered lie detector program is a core AI development involving new algorithms and machine learning.