The Trump administration is considering an executive order that would require a federal review of certain AI models, a move that could directly delay the release schedule of Anthropic, one of the most closely watched companies in the field. Officials have not publicly confirmed the draft, but people familiar with the discussions say the review would target frontier models — the kind Anthropic builds — before they reach the market.
What the order would do
The proposed directive would task agencies like the Commerce Department and the National Institute of Standards and Technology with evaluating large AI models for national security risks, bias, and safety flaws. Companies would need to submit models for review before deployment, a process that could take months. The administration argues the checks are necessary to prevent misuse of powerful AI, but critics say the bureaucracy will choke innovation.
Anthropic, which recently raised billions and is racing to release its next-generation model, would be among the first affected. The company has built its reputation on safety-first development, but a mandatory review cycle would add an unpredictable delay to its roadmap. Anthropic declined to comment on the draft order.
Anthropic in the crosshairs
Anthropic was founded by former OpenAI researchers who left over disagreements on safety practices. The startup has positioned itself as the responsible alternative, releasing models with strict usage policies and third-party audits. But a government review process would shift that dynamic: instead of self-imposed guardrails, the company would face external timelines it can’t control.
The review could cover both the model’s training data and its post-training behavior. That means Anthropic would have to freeze development while the government examines its work. For a company that relies on rapid iteration to stay competitive with OpenAI, Google, and Meta, that’s a major bottleneck.
Broader market and leadership risks
Industry observers warn the order could ripple beyond Anthropic. If the U.S. imposes a review regime that other nations don’t, American companies may fall behind in the global AI race. “This isn’t just about one startup,” said a former White House technology advisor who spoke on condition of anonymity. “It’s about whether the U.S. can still set the pace when other countries are moving without these checks.”
The order would also affect venture capital dynamics. Investors who have poured money into AI labs could see returns pushed further out. Smaller startups without Anthropic’s cash reserves might not survive the waiting period. The administration has not said whether the review would apply retroactively to models already in use, adding another layer of uncertainty.
For now the draft is still being debated inside the White House. No timeline has been set for a formal announcement. What is clear: the clock is ticking for companies like Anthropic, and the next few weeks will determine just how much the government intends to slow them down.
