The Trump Administration Is at War With Itself Over AI Regulation
The White House is split over whether to revive an AI executive order that would let officials review models before release.
Intelligence analysis by GPT-5.4 Mini
Trump’s team is fighting internally over a revised AI regulation plan after canceling an earlier executive order. The main dispute is whether to require or encourage early government access to frontier models for security review.
A group inside the White House is arguing about how to handle powerful AI tools. Some people want the government to look at them early, before companies show them to everyone.
It is a bit like checking a new car for brake problems before it goes on sale. The people who support the idea say it could help spot dangers early. The people against it worry it would slow things down too much.
Nothing is settled yet. The article says the administration is still trying to decide whether to bring the plan back in a new form.
Analysis
What happened
The article says the Trump administration is trying to decide whether it can bring back an executive order on AI regulation after Trump abruptly canceled the original signing. That earlier version was scrapped on May 21, and the central issue was a voluntary framework that would give the White House early access to AI models from labs such as OpenAI, Anthropic, and Google before public release.
Why the order was contentious
According to the piece, the draft order reflected a belief inside the White House that AI has become a national security issue. The article says the models in question can find weaknesses in older software systems, which is part of why officials want to evaluate them earlier. But several executives told WIRED they may not be willing to share models as much as 90 days before release.
Who is pushing and who is resisting
The reporting describes a split inside the administration. White House chief of staff Susie Wiles, treasury secretary Scott Bessent, and national cyber director Sean Cairncross are described as pushing to resurrect the order. Trump’s former AI czar David Sacks is described as opposing it, arguing that the order would be too burdensome and would slow innovation.
Where things stand
The article says Bessent has been meeting with AI executives, including Anthropic CEO Dario Amodei, and may play a role in talks with China on cross-border AI regulation. Other parts of the government, including Commerce and the Pentagon, are said to have played smaller roles or taken different approaches through existing model-testing programs. The clearest takeaway is that the policy is still unsettled, and the final decision depends on whether Trump says yes.
Key points
- The Trump administration is debating whether to revive an AI executive order Trump canceled at the last minute.
- The disputed proposal would have given the White House early access to frontier AI models for security review.
- Susie Wiles, Scott Bessent, and Sean Cairncross are described as pushing to bring it back.
- David Sacks is described as opposing the order and warning it would burden innovation.
- The final decision still depends on whether Trump agrees.
If a revised order survives the internal fight, the White House could get earlier visibility into frontier models and their security risks. The article suggests some officials and executives hope less controversial parts of the proposal could remain intact.
The infighting could leave the administration without a clear AI policy, or stall the order entirely. Companies also may resist sharing models 90 days before release, which could make the framework hard to implement.
