White House advisers reportedly told Meta, Google, OpenAI and Anthropic the government will not safety-test open-weight AI models before release.
Continue to AI University →