Skip to content
Sign in
Defense & Security · North America

DCSA says polygraph modernization excludes generative AI and facial analysis

The Defense Counterintelligence and Security Agency says its Modernizing Polygraph effort does not use generative AI, large language models, facial action coding or micro-expression analysis. The agency made the statement after Defense News asked about concerns tied to earlier Defense Department statements and slides.

AI-assisted synthesis · Based on 2 outlets' reportingHow it's made

View timeline · 7 entries
2 outletsLatest reporting 3 min read
Coverage: Too few ratedL 0 · C 2 · R 0
Updat3 graphic: DCSA
Updat3 graphic

What happened

Key facts

  • The Defense Counterintelligence and Security Agency said its Modernizing Polygraph effort does not incorporate generative AI (per c4isrnet.com).
  • The agency said its research does not involve facial action coding or micro-expression analysis (per c4isrnet.com).

The brief

The Defense Counterintelligence and Security Agency says its Modernizing Polygraph effort excludes generative artificial intelligence, large language models, facial action coding and micro-expression analysis.

The agency gave that account after Defense News asked about concerns stemming from earlier Defense Department statements and slides describing AI tools for assessing staff trustworthiness. The agency’s statement drew a distinction between the polygraph modernization effort and the broader set of tools described in earlier Defense Department material.

Those earlier statements and slides referred to artificial intelligence systems analyzing text, voices and faces to estimate trustworthiness. The agency’s current account says the research does not use the named generative or facial-analysis methods; it does not, by itself, explain how the earlier descriptions relate to the effort.

The Defense Counterintelligence and Security Agency vets security clearance holders and applicants. That role makes the stated boundaries of a modernization project relevant to people seeking or holding clearances, though the coverage does not describe a change to clearance decisions or identify anyone affected by the tools.

The earlier material described AI-based assessment as support for investigations. A quoted statement characterized such analysis as a “force multiplier” for human investigators, rather than a replacement.

The agency’s more recent description, by contrast, specifies categories it says the Modernizing Polygraph effort does not incorporate, including chatbots and methods based on facial movement. A speaker quoted in the coverage invoked the 2002 film Minority Report, which depicts a government using psychic children to anticipate crimes and detain people accused of future offenses.

That comparison conveys concern about predictive assessment, but the reported agency statement concerns methods it says are excluded from the current research effort, not a claim that the project predicts crimes.

The central issue is the relationship between the agency’s description of the project and the earlier Defense Department descriptions of trustworthiness-analysis tools. The agency’s statement followed a question about those earlier materials, while the coverage does not provide a detailed account reconciling the two descriptions.

The reporting therefore identifies a difference in scope and emphasis, not a confirmed finding that the agency is using the excluded technologies. For clearance holders and applicants, the practical question is what methods the agency uses in the modernization effort and how it defines the boundaries it described.

The immediate record is the agency’s stated exclusions alongside prior Defense Department descriptions of AI analysis; no specific operational deployment or effect on individual clearance cases is reported.

Why it matters

  • Security clearance holders and applicants are the population most directly connected to the agency’s role: the Defense Counterintelligence and Security Agency vets them, and the earlier material described AI analysis intended to estimate staff trustworthiness.
  • The specific mechanisms at issue are generative AI, large language models, facial action coding and micro-expression analysis, which the agency says its Modernizing Polygraph effort excludes.
  • Earlier Defense Department statements and slides described analysis of text, voices and faces; the contrast leaves the scope of the modernization effort as a practical concern for people subject to clearance vetting.

What to watch next

  • Whether the Defense Counterintelligence and Security Agency further explains how its project description relates to the earlier Defense Department statements and slides.
  • Whether the agency gives more detail about what methods its Modernizing Polygraph effort does use.
  • Whether the Defense Department clarifies the relationship between trustworthiness-analysis tools described earlier and the agency’s stated exclusions.

How it's being covered

Coverage analysisShow details
2 source articles · 2 unique outlets
Outlet political mix: Too few rated· 2 Center
Where the outlets are based: US/Canada 50% · Unknown 50%
Outlet political mix
Too few rated
Left
Center
Right
0 outlets
2 outlets
0 outlets
The marker shows coverage balance: each unique outlet counts once.
Geography
Where the outlets covering this story are based.
2 unique outletsDominant coverage: US/Canada
Limited coverage
  • Coverage is currently limited to 2 outlets.
How we measure coverage

How outlets' accounts compare

2 framing differences · 2 uncorroborated claims · 3 information gaps
The reports repeat the same account with different emphasis; its central claims are not independently confirmed in the reporting Updat3 read.Compared across 2 reports whose article text Updat3 read.
Framing / emphasis differences
Sources emphasise different angles, actors or consequences of the same reporting.
  • The Defense Counterintelligence and Security Agency’s account focuses on methods it says the Modernizing Polygraph effort excludes; earlier Defense Department statements and slides described AI tools analyzing text, voices and faces to estimate staff trustworthiness.
  • The agency said the Modernizing Polygraph effort excludes the named technologies; earlier Defense Department statements and slides described AI tools for estimating staff trustworthiness.
Updat3: These are differences of emphasis, not disputes about what is reported. They do not show that the reported claims are true.
One-sided or uncorroborated claims
Claims that come from one party or one source family and are not independently corroborated.
  • A quoted statement called AI-based analysis a “force multiplier” for human investigators, not a replacement.
  • A speaker quoted in the coverage compared predictive assessment to Minority Report.
Updat3: Not independently corroborated yet — treat as a claim, not an established fact.
Not established by the reporting
Information the available reporting does not contain or confirm. This is a gap, not a disagreement.
  • The reporting does not specify which methods the Modernizing Polygraph effort does use.
  • The reporting does not describe any operational deployment of the tools or effects on individual clearance decisions.
  • The relationship between the agency’s stated exclusions and the earlier Defense Department descriptions is not reconciled in the reporting.
Updat3: Updat3 cannot conclude anything about this from the current sources.

Timeline

Developments from this story's sources, dated background we compiled and fact-checked, and our earlier coverage. Each row says which.
  1. John Augustus Larson developed an instrument that recorded multiple physiological signals during questioning.
    Background · compiled and fact-checked by Updat3
  2. A federal appeals court rejected polygraph evidence in Frye v. United States, finding its underlying principle had not gained general acceptance.
    Background · compiled and fact-checked by Updat3
  3. Congress enacted the Employee Polygraph Protection Act, barring most private employers from requiring lie-detector tests.
    Background · compiled and fact-checked by Updat3
  4. President Bill Clinton signed Executive Order 12968, setting standards for access to classified information and allowing polygraphs in specified security contexts.
    Background · compiled and fact-checked by Updat3
  5. In *United States v. Scheffer* on March 31, 1998, the U.S. Supreme Court upheld a military rule excluding polygraph evidence from courts-martial, leaving military judges discretion over its admission.
    Background · from our historical context
  6. The National Research Council published The Polygraph and Lie Detection, examining the scientific evidence behind polygraph use.
    Background · compiled and fact-checked by Updat3
  7. The Defense Counterintelligence and Security Agency was established to handle personnel security, industrial security and related missions.
    Background · compiled and fact-checked by Updat3

How we got here

The most consequential earlier review of the science behind polygraph use was the National Research Council’s 2003 report, *The Polygraph and Lie Detection*. The Employee Polygraph Protection Act of 1988 barred most private employers from requiring lie-detector tests, while exempting federal, state and local governments.

Executive Order 12968, signed by President Bill Clinton on August 2, 1995, set government-wide standards for access to classified information and allowed agencies to use polygraph examinations in specified security contexts. The 2003 National Research Council report examined the limits of polygraph evidence within this broader history of government use.

Original sources

2 articles from 2 outlets

Every story links to the reporting it was built from. How we choose sources

Spotted a factual error, a broken link or a misleading summary? Tell us.