The Pentagon is using polygraphs to hunt leakers, but the tests are more theater than fact

In August 2026, government investigators subjected roughly 50 military officers and civilian employees on the Joint Staff to polygraph tests as part of an inquiry into press leaks about the Iran war and depleted U.S. munitions, according to The New York Times. The newspaper described the scope of the investigation as unprecedented in the modern military.

This is not the first time the Trump administration has used polygraphs in this way; in 2025, it directed that a wave of tests be administered across the executive branch to expose minor leaks.

As someone who has taken roughly a dozen polygraphs during my 27-year career with the CIA, I am wary of how the process is understood – and of the certainty its use can appear to convey.

Polygraphs carry an ominous, almost mythological reputation among Americans. The more familiar and unofficial term – lie detector tests – likely fuels that perception. Television crime dramas have done their part, too, often portraying the device as an oracle for uncovering the truth when conventional methods fail.

In those portrayals, the polygraph is not merely a tool – it’s a window into the soul.

Among those entering government service, especially in national security, the greater anxiety is not the background check but passing the polygraph. My advice is always the same: Don’t lie. It’s the best – and perhaps only – guidance for a process that most assessments have concluded is a more subjective interpretation than empirical science.

Polygraphs are “pseudo-scientific” in that they measure physiological responses such as heart rate, blood pressure and perspiration. The assumption is that liars betray themselves through spikes in those signals. But this presumes a kind of psychological transparency that simply doesn’t hold up. A person might sweat and tremble simply from fear, anger or frustration – not deceit.

There also are no specific physiological reactions associated with lying. The National Academy of Sciences in 2003, and the American Psychological Association in a 2004 review, concluded that the polygraph rests more on theater than fact. A 2019 assessment reached the same conclusion.

Accordingly, polygraph results are not generally admissible in U.S. courts. Only a handful of states – such as Georgia, Arizona and California – permit their use even under limited conditions. And they typically require that both parties agree to admission and a judge to approve it. Unconditional admissibility remains the exception, not the rule.

And yet, inside many national security agencies, polygraphs remain central to the clearance process – a fact I observed firsthand during my time overseeing personnel vetting and analytic hiring within the intelligence community.

While not treated as conclusive, polygraph results often serve as a filter. A candidate’s visible discomfort – or the examiner’s subjective judgment that a response seems evasive – can stall or end the hiring process. For instance, I know that government agencies have halted clearances after an examiner flagged elevated reactions to questions about past drug use or foreign contacts, even when no disqualifying behavior was ultimately documented.

In some cases, an examiner’s suggestion that a chart shows an anomaly has led otherwise strong applicants to volunteer details they hadn’t planned to share – such as minor security infractions, undeclared relationships, or casual drug use from decades earlier – that, while not disqualifying on their own, reshape how their trustworthiness is perceived.

The polygraph’s power lies in creating the conditions under which deception is confessed.

No administration has been immune to the impulse to investigate leaks. The reflex is bipartisan and familiar: An embarrassing disclosure appears in the press – contradicting official statements or exposing internal dissent – and the White House vows to identify and punish the source. Polygraphs are often part of this ritual.

During his first term, Trump intensified efforts to expose internal dissent and media leaks. Department guidelines were revised to make it easier for agencies to obtain journalists’ phone and email records, and polygraphs were reportedly used to pressure officials suspected of talking to the press. That trend has continued – and, since 2025, escalated.

Policies at the Pentagon now restrict unescorted press access, revoke office space for major outlets and favor ideologically aligned networks. The line between legitimate leak prevention and the surveillance or sidelining of critical press coverage has grown increasingly blurred.

At agencies such as the Department of Homeland Security and the FBI, polygraphs reportedly have been used frequently – and more punitively – to identify internal dissenters. Even “cold cases,” such as the leak of the Supreme Court’s Dobbs opinion ahead of its overturning of Roe v. Wade, were reopened, despite prior investigations yielding no definitive source.

Not all leaks are treated the same. Disclosures that align with official narratives or offer strategic advantage may be quietly tolerated, even if unauthorized. Others, especially those that embarrass senior officials or reveal dysfunction, are more likely to prompt formal investigation.

In 2003, for example, the leak of CIA officer Valerie Plame’s identity – widely seen as retaliation for her husband’s criticism of the Iraq War – triggered a federal investigation. The disclosure embarrassed senior officials, led to White House aide Scooter Libby’s conviction for perjury, later commuted, and drew intense political scrutiny.

Leaks involving classified material draw the sharpest response when they challenge presidential authority or expose internal disputes. That was the case in 2010 with Chelsea Manning, whose disclosure of diplomatic cables and battlefield reports embarrassed senior officials and sparked global backlash. Government reaction often depends less on what was disclosed than on who disclosed it – and to what effect.

A narrow set of disclosures, such as those involving espionage or operational compromise, elicit broad consensus as grounds for prosecution. But most leaks fall outside that category. Most investigations fade quietly. The public rarely learns what became of them. Occasionally, there is a vague resignation, but direct accountability is rare.

The Pentagon polygraph campaign is not likely to eliminate leaks to the press. But it may have a chilling effect that discourages internal candor while diverting investigative energy away from core security priorities.

Even if such campaigns succeed in reducing unauthorized disclosures, they may come at the cost of institutional resilience. Historically, aggressive internal enforcement has been associated with declining morale and reduced information flow – factors that can hinder adaptation to complex threats.

Some researchers have suggested that artificial intelligence may eventually offer reliable tools for detecting deception but caution that the technology is nowhere near operational readiness.

For now, institutions will have to contend with the tools they have – imperfect, imprecise and more performative than predictive.

This is an updated version of a story that was first published on July 9, 2025.

This article is republished from The Conversation, a nonprofit, independent news organization bringing you facts and trustworthy analysis to help you make sense of our complex world. It was written by: Brian O'Neill, Georgia Institute of Technology

Brian O'Neill does not work for, consult, own shares in or receive funding from any company or organization that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.