The Department of Commerce's Center for AI Standards and Innovation (CAISI) has initiated agreements with major AI firms, including Microsoft, xAI, and Google, allowing governmental oversight of unreleased AI models. This development follows earlier agreements made with Anthropic and OpenAI in 2024.
On May 5, 2026, CAISI announced that these partnerships aim to enhance evaluations of advanced AI systems and improve AI security. The collaborations are part of efforts to align with the directives of the secretary of commerce and the broader AI Action Plan of the United States.
However, the official announcement is currently unavailable on the CAISI website, leading to an error page when accessed. Reports indicate that the link now redirects to the main CAISI page. As of Monday evening, no response has been received from the White House or the Commerce Department regarding the missing information.