Only Congress can put meaningful limits on government abuse of AI.

The deal comes as rival Anthropic is locked in a more contentious standoff with U.S. defense officials.

OpenAI CEO Sam Altman said on X that his company has come to terms with the Defense Department for use of its models.