US Secures Early Access to Frontier AI Models from Google, Microsoft, xAI for Security Testing

The U.S. Department of Commerce has reached new agreements with leading AI developers including Google DeepMind, Microsoft, and xAI to allow early government access to advanced artificial intelligence models for national security evaluation.

Announced Tuesday, the initiative will be coordinated through the department’s Center for AI Standards and Innovation. The partnerships are designed to enable pre-deployment testing and research, helping authorities better understand the capabilities, limitations, and potential risks of cutting-edge AI systems before they are released to the public.

Under the agreements, the government will also conduct post-deployment assessments and collaborate with developers on improving safety practices. Officials say the program strengthens oversight at a time when AI technologies are rapidly advancing.

According to the center, more than 40 evaluations have already been completed, including assessments of unreleased frontier models. In some cases, developers provide versions of their systems with reduced safeguards to allow deeper analysis of potential national security implications.

“Independent, rigorous measurement science is essential to understanding frontier AI and its national security implications,” said Director Chris Fall, noting that the expanded collaborations will help scale evaluation efforts “at a critical moment.”

The Center for AI Standards and Innovation has been designated as the federal government’s primary hub for industry engagement on AI testing, joint research, and the development of best practices for commercial AI systems.