Google, Microsoft, and xAI have signed agreements with the Commerce Department to provide early versions of their AI models to the government for safety and capability assessments before public release. This brings all five major US AI firms — including OpenAI and Anthropic, who signed on in 2024 — under a unified pre-release review framework.
The evaluations are run by the Center for AI Standards and Innovation, a Commerce Department agency that began as the Biden administration's AI Safety Institute in 2023 before being renamed and restructured last year under Trump. The center has already completed more than 40 model assessments, including some on cutting-edge systems not yet released to the public. Existing agreements with OpenAI and Anthropic have also been renegotiated to align with the Trump AI action plan.
Center director Chris Fall said the new deals will help "expand work on public interest at a critical moment." He took over from Collin Burns, a former Anthropic AI researcher who was forced out just days after taking the role.
Before the agreements were announced, both The New York Times and The Wall Street Journal reported that the Trump administration was considering an executive order to create a formal government review process for AI tools. White House officials said those discussions were still speculative and that any announcement would come directly from the president.