What is the US government's Polygraph+ program, and what new technologies will it use?

The US government's Polygraph+ initiative is a $30.3 million program aiming to modernize federal lie detection over five years. Run by the Defense Counterintelligence and Security Agency, it will integrate artificial intelligence and machine learning for scoring algorithms. A key innovation is "standoff sensing," allowing physiological measurements like heart and breathing rates to be taken without direct physical contact. The goal is to enhance accuracy and reliability for employee vetting and insider threat detection.

How accurate are traditional polygraph tests, and why do experts generally doubt their reliability?

Traditional polygraph tests measure physiological changes like blood pressure, pulse, breathing, and sweat to infer deception. However, their reliability is widely questioned by experts and government studies, including the 2003 US National Research Council report, which found evidence of efficacy "weak at best." Results are rarely admissible in court, and tests can be inaccurate, subjective, and vulnerable to countermeasures, as there is no single, universally true physiological sign of lying.

Will artificial intelligence reliably improve future lie detection accuracy, or are there still significant challenges?

While artificial intelligence could theoretically improve lie detection by identifying complex patterns and combining multiple physiological cues, significant challenges persist. Critics argue AI adds uncertainty to an already invalid methodology, as there's no reliable "ground truth" for what constitutes a lie to train algorithms effectively. Previous AI-enhanced lie detection efforts have not yielded consistently reliable results outside laboratory settings, raising doubts about its practical application for accurate deception assessment.

← Back to front page

AI Breakthroughs & Applied ResearchFriday, September 25, 2026

The Pentagon wants $30 million to build an AI-powered lie detector

Original reporting by MIT Technology Review

Polygraph+ (or Polygraph Next) refers to a $30.3 million US government initiative to develop an advanced lie detection system, integrating artificial intelligence with non-contact physiological sensing. Spearheaded by the Defense Counterintelligence and Security Agency, this five-year program aims to modernize federal polygraph technologies, enhancing accuracy and reliability for vetting prospective employees and identifying insider threats. The investment arrives amidst heightened tensions within the Department of Defense, which has recently intensified its use of traditional polygraph tests to uncover sources of alleged leaks. The proposed system will leverage AI and machine learning for sophisticated scoring algorithms and explore "standoff sensing," allowing physiological measurements without physical contact.

A history of doubt

Yet, this ambitious project encounters a formidable history of scientific skepticism. Traditional polygraphs, largely unchanged since their 1920s invention, have been repeatedly debunked by scientific bodies, with their results rarely admissible in court. Experts caution that while AI might identify new patterns in data, it cannot reliably link them to deception in the absence of a verifiable "ground truth." Critics fear that Polygraph+ represents a "misguided effort" that combines AI's complexity with a fundamentally invalid premise, potentially creating a system used more as a psychological prop to elicit confessions than a reliable arbiter of truth.

The Department of Defense’s renewed push for an AI-powered polygraph, dubbed “Polygraph+,” represents the latest chapter in a long-standing, often fraught, quest for technological truth detection. Despite significant investment and the promise of advanced machine learning and standoff sensing, the initiative confronts the same fundamental hurdle that has stymied all previous efforts: the absence of a reliable, universal physiological marker for deception. Experts caution that integrating AI into an already scientifically unsound methodology risks amplifying its flaws, creating a more sophisticated but equally unreliable system susceptible to misuse and lacking crucial "ground truth" for validation.

Broader Implications

This endeavor extends beyond mere technological development; it reflects a deeper organizational imperative. Against a backdrop of heightened concerns over leaks and perceived lack of loyalty, Polygraph+ appears less a scientific breakthrough and more a strategic tool, intended to deter and intimidate rather than definitively uncover truth. Should this technology be deployed, even with its inherent uncertainties, it carries substantial implications for civil liberties and due process, particularly for federal employees. The potential for false accusations, driven by a system leveraging unproven AI on an unproven premise and applied at scale, could erode trust, disproportionately impact minority groups, and create a chilling effect across the public sector. Ultimately, without addressing the core scientific limitations of deception detection, Polygraph+ risks becoming another expensive, technologically advanced solution to a problem that remains fundamentally unresolved, underscoring the persistent challenge of distinguishing genuine scientific progress from the alluring but dangerous promise of technological panaceas.

Frequently asked questions

What is the US government's Polygraph+ program, and what new technologies will it use?: The US government's Polygraph+ initiative is a $30.3 million program aiming to modernize federal lie detection over five years. Run by the Defense Counterintelligence and Security Agency, it will integrate artificial intelligence and machine learning for scoring algorithms. A key innovation is "standoff sensing," allowing physiological measurements like heart and breathing rates to be taken without direct physical contact. The goal is to enhance accuracy and reliability for employee vetting and insider threat detection.
How accurate are traditional polygraph tests, and why do experts generally doubt their reliability?: Traditional polygraph tests measure physiological changes like blood pressure, pulse, breathing, and sweat to infer deception. However, their reliability is widely questioned by experts and government studies, including the 2003 US National Research Council report, which found evidence of efficacy "weak at best." Results are rarely admissible in court, and tests can be inaccurate, subjective, and vulnerable to countermeasures, as there is no single, universally true physiological sign of lying.
Will artificial intelligence reliably improve future lie detection accuracy, or are there still significant challenges?: While artificial intelligence could theoretically improve lie detection by identifying complex patterns and combining multiple physiological cues, significant challenges persist. Critics argue AI adds uncertainty to an already invalid methodology, as there's no reliable "ground truth" for what constitutes a lie to train algorithms effectively. Previous AI-enhanced lie detection efforts have not yielded consistently reliable results outside laboratory settings, raising doubts about its practical application for accurate deception assessment.

Intro and outro generated by Printing Press AI from the source article above. Always consult the original reporting for verbatim quotes and primary sources.