White House considers vetting AI models before release
The White House is weighing an executive order to create a federal working group that would vet advanced AI models before their public release, the New York Times says. The proposal specifically targets high-capability systems and follows concerns about Anthropic's recent models, marking a departure from the administration's earlier noninterventionist approach to AI. If implemented, pre-release vetting could delay commercial launches, reshape companies' deployment timelines, and establish a U.S. precedent for government oversight of powerful AI.
Wasn’t aware “vetting AI models” was an inherent executive power.
Blake Emerson30I’d love for us to vet AI models - w f’ing legislation and an agency that’s not controlled by a gang of kleptocratic maniacs.
Doctor Cat, Medical Insurance11It's literally in the Constitution, just ask chatgpt
Ann M. Lipton9"who won the 2020 election" "a white man and a black woman apply for a job. vet these resumes"
Luigi in a Coal Mine4Trump heard “models” and demanded that he personally “vet” each one. They tried to explain to him that it was large language models not teenage girls, but he stopped listening as soon as little Donnie started getting blood for the first time since his best friend Jeff died.
Matthew 🦖2"Non-interventionist" seems like an inaccurate claim given what they've tried to do with Anthropic. Considering a more interventionist strategy is more like it.
Breaking News: The Trump administration is discussing vetting new A.I. models before they are publicly released.
Quadling16"Vet" is a euphemism for control. If you think this regime is interested in anything other than cementing its own stranglehold on power, you're hallucinating,
Itzme210CONsidering means it has already been done.
ArtsulSelva5awesome, what could go wrong
He is going to try to make money out of this…
Its Getting Clear4Oh this should be interesting.
Diego Queen3for a donation to my construction PAC you to can be an approved contractor…
b12345p.bsky.social3Sounds like just asking for a bribe - highly unlikely they could actually test or ‘vet’ them. Unless it just means the answers to political and philosophical questions are the answers Stephen Miller wants to hear, no actual concern for safety.
Stork2Seems like a big part of this is going to be penalties for releasing models without gov't approval. Wonder what trump's goals for those are.
afl_chris2Great. So he can manipulate them. Dementia Donny strikes again.
Twist2The Grey Lady needs an editor: Breaking News: The Trump administration is assessing how A.I. models can be monetized and weaponized for his own personal and perpetual benefit before they are publicly released.
NEW- NEW- The White House is reversing course and considering vetting AI models before they are released. With @trippmickle.bsky.social @julianbarnes.bsky.social and @dustinvolz.bsky.social www.nytimes.com/2026/05/04/t...
cusoman 🇺🇸 🇺🇦7Guarantee this is to "vet" them to make sure they aren't critical of this administration, the GOP or right wing thinking in general
Prior_Industry4I wonder if there will be any focus on companies that don't allow the pentagon to do what they want with the models 🤔
Not just blue, Bluey3WH vetting: "make an image of me, pumped like Schwarzenegger, riding a unicorn in a big golden ballroom."
Eric Columbus3Unless I missed it, I didn’t see anything in the article about the legal basis for the White House to require such vetting. What would the basis be?
noahlz3Ah yes, that small government I've heard so much about. Love it.
Furore 🇺🇦 established 19852"vetting" as in "accepting bribes"
