The US Department of Defense has requested $30.3 million over the next five years to develop an advanced lie-detection system referred to in budget documents as “Polygraph+” or “Polygraph Next.” The program will emphasize AI- and machine learning-driven scoring algorithms and a method known as “standoff sensing,” which means taking physiological measurements from a subject without attaching sensors to their body.
According to the budget request, first reported by Inside Defense, the aim is to "modernize federal polygraph and credibility assessment technologies" to increase accuracy and reliability. The program would be run by the Defense Counterintelligence and Security Agency (DCSA), which handles background checks for the federal government. The document says the technology would be used for vetting prospective employees and for "insider threat detection." The budget request has not yet been approved by Congress, and the DCSA did not respond to requests for more details.
The announcement comes amid heightened internal tensions at the Department of Defense. Under Defense Secretary Pete Hegseth, the Pentagon has increasingly used polygraph tests to try to identify sources of alleged leaks to the press. The New York Times reported in September that around 50 officers on the Joint Staff had been given polygraph tests after media coverage about depleted US weapons stockpiles in the war with Iran.
Which specific technologies will be incorporated into Polygraph+ is not yet clear. However, prior DoD efforts offer possible clues: in 2023 the Defense Innovation Unit (DIU) ran an open call to find companies with deception-detection products. The DIU selected two companies to build prototypes: Presage Technologies, which says it can measure heart rate and breathing using standard cameras, and Altec Research, a medical sensor company moving into non-contact sensing.
A screenshot of Altec’s prototype released by the DIU shows the system tracking head movement, facial skin temperature and pore activity. Presage Technologies and Altec Research did not respond to requests for comment, and the DIU declined to comment.
Traditional polygraph technology has changed little since its invention in the 1920s. Examiners measure blood pressure, pulse, breathing and sweat and determine veracity by comparing physiological responses to baseline questions (for example, "Is the sky blue?") and relevant questions (for example, "Have you ever committed a crime?").
The federal government conducts tens of thousands of polygraph tests annually as part of employee screening, yet the polygraph has repeatedly been debunked and its results are rarely admissible in court. In 1983, the Congress’s Office of Technology Assessment concluded that evidence supporting the polygraph’s use for employee screening was very limited; in 2003 the US National Research Council (NRC) determined that evidence of efficacy was "weak at best."
Research indicates people can spot a lie just over half the time without technical aids. The American Polygraph Association claims the polygraph is 80–94% accurate, but the 2003 NRC report noted that screening tests at that level of accuracy could still produce many errors. With the Department of Defense employing about 2.8 million staff, an imperfect system at scale could falsely accuse tens of thousands of people.
There are further problems: polygraph interpretations can be subjective, with different examiners producing widely different results, and people from minority groups are more likely to be judged deceptive. Interviewees can also learn countermeasures—such as artificially increasing physiological responses to baseline questions by stepping on a hidden pin in their shoe—to try to beat the test.
"If you know how it works, you can beat it," says Sophie van der Zee, an associate professor at Erasmus University who studies deception. She argues the instrument’s biggest effect is deterrence: subjects often confess before the test even begins. "But that only works if people think a polygraph works."
Over the decades there have been numerous attempts to develop new lie-detection technologies—from thermal cameras to pupil trackers to brain scans—but none have produced reliable results outside laboratory settings. The structural problem, experts say, is that there is no single universal sign of lying true for everyone at all times. "There is still no Pinocchio’s nose," van der Zee says.
AI could theoretically improve polygraphs if it identifies patterns in data that human examiners miss. Machine-learning approaches lend themselves to "multi-modal" deception detection, combining multiple measurements into an overall deception "score" that is harder to game. Van der Zee says deception involves three underlying phenomena that detection tries to pick up on: physiological stress, cognitive load, and conscious efforts to conceal lying. Current polygraph technology typically addresses only the first.
Some critics caution that combining AI with polygraph methods compounds problems rather than solving them. Kyri Kotsoglou, a professor at Northumbria Law School who studies polygraph use in the justice system, calls the effort "misguided" because it attempts to make the complex tangible. Marion Oswald, a law professor who has written with Kotsoglou on polygraph use, warns there is no reliable ground truth: even extensive polygraph records cannot confirm whether past polygraph results were correct.
Oswald fears new lie-detection forms will, like the traditional polygraph, be used more as psychological props than as scientific tools. "It seems very much a response to the concern of the current administration to leaks and perceived lack of loyalty," she says. "[Lie detection is] being used as a threat, to intimidate and force people to confess to things, as opposed to anything that’s actually getting valid information."
In sum, Polygraph+ aims to modernize and potentially improve lie-detection using AI and contactless sensors, but decades of scientific criticism, questions about fairness and subjectivity, and concerns about misuse leave the program’s potential benefits and risks uncertain.
