Free daily briefing on global business news.
The framework, ordered by a June executive order, is expected to target closed-model companies OpenAI, Anthropic, and Google
Anadolu / Getty Images
The Trump administration told leading artificial intelligence companies on Tuesday that open-weight AI models would be exempt from a new government security review framework, focusing scrutiny instead on closed, proprietary systems from companies like OpenAI, Anthropic, and Alphabet $GOOGL's Google.
According to The Wall Street Journal, the framework would require pre-release government testing only from developers of closed, proprietary models whose systems score at the frontier on cybersecurity and hacking evaluations. Models released as open weights, giving anyone the ability to run and adapt them locally, would fall outside the review requirement. Nvidia $NVDA and other open-weight model makers could eventually be required to submit their tools for testing as they grow more capable.
