US Commerce Dept. quietly pulls page on AI security deals with Google, Microsoft, xAI
Pre-deployment testing agreements with three tech giants vanish without explanation...
On May 12, 2026, reports emerged that the US Commerce Department had quietly removed a webpage detailing new AI security testing agreements with Google DeepMind, Microsoft, and xAI. The page, originally published on May 5, 2026, announced pre-deployment evaluations by the Center for AI Standards and Innovation (CAISI). Now, visitors either receive an error message or are redirected to the CAISI homepage. The reason for this removal remains unexplained, sparking speculation and concern among AI governance observers.
The removed webpage was part of a broader initiative to establish voluntary safety testing for frontier AI models before public release. The involvement of major players like Google DeepMind, Microsoft, and xAI suggested a collaborative approach to AI risk mitigation. The sudden withdrawal of this information without any official statement raises questions about the status of these agreements and the transparency of AI oversight efforts. Industry watchers are now calling for clarity on whether the testing framework has been suspended, altered, or simply moved behind closed doors.
- The Commerce Department removed a May 5, 2026 page detailing AI security testing pacts with Google DeepMind, Microsoft, and xAI.
- The page announced pre-deployment evaluations by the Center for AI Standards and Innovation (CAISI), now redirecting to its homepage.
- No official explanation was given for the removal, leaving the status of the agreements unknown.
Why It Matters
Unexplained removal of public AI safety testing agreements could signal reduced transparency in federal AI oversight.