The US government has removed information from an official government website related to an agreement with tech companies Google, Microsoft, and xAI. The removal was reported in May 2026, though the reason behind it remains unclear.
The deleted content concerned a pact between the US government and the three tech giants that allowed government scientists to test new AI models for security flaws before those models were released to the public. The agreement was designed to identify potential vulnerabilities in advanced AI systems ahead of their public deployment.
The removal comes amid growing concerns about national security risks posed by advanced AI systems, though officials have not publicly explained why the information was taken down.
The disappearance of this content may raise questions about the current status of the testing arrangement and the government’s approach to AI security oversight, though no official statement has confirmed any change to the underlying agreement itself.
Source: Tech-Economic Times