[ad_1]
In early October 2025, a rumor began circulating on social media claiming that U.S. President Donald Trump used artificial intelligence and facial-recognition technology to assess the loyalty of U.S. generals at a meeting the president and Defense Secretary Pete Hegseth gathered in Quantico, Virginia, on Sept. 30.
One Facebook post (archived) with the claim read:
SoftwareDefinedGovernance:
“The reason Trump and Pete Hegseth gathered all those generals in one room was not just about a loyalty speech or a pep rally. I’m hearing that the Trump team used artificial intelligence and facial-recognition technology during those briefings to monitor the generals’ reactions in real time. Every eyebrow raise, every flicker of doubt, every moment of discomfort was scanned and analyzed by an algorithm designed to detect who would obey orders without question—and who might resist.”
(Facebook user Alexander Ainslie)
Similar versions of the claim appeared across Reddit, Facebook, X and online blogs.
Snopes has opted not to rate this claim because we could not definitively disprove it. That said, there was no evidence to support the claim that Trump used AI or facial recognition technology to scan U.S. generals for signs of loyalty or disloyalty.
Searches across multiple search engines revealed no credible reporting to suggest that such an operation ever took place and the rumor spread primarily via social media. Given the extraordinary nature of the allegation, which implied the use of surveillance technology on high-ranking military officials within a sensitive national security context, it is highly unlikely that such a program could have been implemented without drawing significant attention from reputable media outlets, watchdog organizations or whistleblowers. Furthermore, the mechanics of how AI would interpret facial expressions to determine “loyalty” were not grounded in any known methodology.
We reached out to the White House to ask if Trump ever undertook any such initiative involving facial recognition or AI in this context and will update this article if we receive a response.
Source of the claim
The story originated from an Oct. 3 Substack post (archived) by Lev Parnas titled “Breaking: Trump Used AI to Scan U.S. Generals’ Faces for Loyalty — and to Root Out Whistleblowers.” Parnas, an American businessman and former associate of Trump ally Rudy Giuliani, was convicted in 2022 of financial crimes related to Trump’s 2016 election campaign.
We contacted Parnas to ask about the source of the information he shared in his Substack post but did not receive a response as of this writing.
The relevant part read:
The reason Trump and Pete Hegseth gathered all those generals in one room was not just about a loyalty speech or a pep rally. I’m hearing that the Trump team used artificial intelligence and facial-recognition technology during those briefings to monitor the generals’ reactions in real time. Every eyebrow raise, every flicker of doubt, every moment of discomfort was scanned and analyzed by an algorithm designed to detect who would obey orders without question—and who might resist.
Parnas did not provide any sources or verifiable evidence to support this claim, relying solely on the vague phrase “I’m hearing,” without naming where the information allegedly came from. Additionally, there have been no credible reports indicating that any such technology was used, and it remains unclear how AI or facial recognition could reliably assess the “loyalty” of generals.
The post continued:
I’m also hearing from my sources that this isn’t limited to the generals. This same technology is being used as a tool inside Trump’s orbit — a quiet weapon of power, deployed by his loyalists to identify and weed out whistleblowers and anyone who isn’t completely obedient.
Parnas further described the alleged program as a “quiet introduction of a new loyalty test: not just pledging allegiance to Trump, but being scanned and scored by AI for obedience,” claiming the administration was deploying it “against the highest levels of America’s own military leadership.”
The article circulated alongside what appeared to be an AI-generated image purporting to show Trump standing next to a group of U.S. generals. An analysis using Hive, an AI-detection tool, determined with 100% confidence that the image was likely AI-generated. Similarly, the detection tool Sightengine assessed the image as likely AI-generated, with 99% confidence. (Research shows AI-detection software is imperfect. Readers should consider the tools’ results with skepticism.)

(sightengine.com, Hive Moderation)
In October 2025, we also fact-checked whether Hegseth said: “It’s tiring to look out at combat formations, or really any formation, and see fat troops. Likewise, it’s completely unacceptable to see fat generals and admirals in the halls of the Pentagon and leading commands around the country and the world.”
Sources
“AI Image Detector. Detect AI-Generated Media at Scale.” Sightengine, https://sightengine.com/detect-ai-generated-images. Accessed 6 Oct. 2025.
FiscalNote, Roll Call. “Roll Call Factba.Se – Speech: Donald Trump Addresses Military Leadership in Quantico, Virginia – September 30, 2025.” Roll Call Factba.Se, https://rollcall.com/factbase/trump/transcript/donald-trump-speech-department-of-defense-leaders-quantico-september-30-2025. Accessed 6 Oct. 2025.
Hive Moderation. https://hivemoderation.com/. Accessed 6 Oct. 2025.
Ibrahim, Nur. “Real Hegseth Quote about ‘fat Generals and Admirals’ Spreads Online.” Snopes, 2 Oct. 2025, https://www.snopes.com//fact-check/pete-hegseth-fat-generals-admirals/.
Parnas, Lev. “Breaking: Trump Used AI to Scan U.S. Generals’ Faces for Loyalty — and to Root Out Whistleblowers.” Substack newsletter. Lev Remembers, 3 Oct. 2025, https://levremembers.substack.com/p/breaking-trump-used-ai-to-scan-us.
Stewart, Phil, et al. “Hegseth Slams ‘fat Generals,’ Trump Touts Cities as Troop ‘Training Grounds.’” Reuters, 1 Oct. 2025. United States. www.reuters.com, https://www.reuters.com/world/us/trump-preside-over-unusual-military-gathering-virginia-2025-09-30/.
[ad_2]
Aleksandra Wrona
Source link