Jensen Huang Smiles as US Backs Off Banning Chinese Open-Weight AI

Jensen Huang is smiling. The White House has reportedly decided not to go after Chinese open-weight AI models — at least not through its new AI safety review framework.

In a closed-door meeting on August 4, the White House told OpenAI, Google, Anthropic, and other companies that open-weight models will be left out of the new AI safety review framework. That means Chinese models like Kimi K3 and DeepSeek won't have to be handed over to the US government for testing before release. For now, at least, the US isn't using this framework to ban them.

The rise of Kimi K3 sparked fierce debate in Washington over whether to restrict Chinese open-weight models. Some officials considered trade blacklists, security advisories, and even an executive order aimed specifically at open models. Anthropic CEO Dario Amodei has also argued that open models...