Nvidia CEO says US firms can use Chinese AI models

Nvidia CEO Jensen Huang has publicly advocated for American companies to have access to Chinese open-source artificial intelligence models, positioning himself against U.S. officials who are considering restrictions on national security grounds. Huang made the remarks in an interview with Axios on Tuesday, July 21, 2026, as the release of a powerful new Chinese model, Kimi K3, has stirred significant concern in Washington and on Wall Street.

Huang argued that competition between open and closed-source models is positive for the global AI industry and that excellent open-source models should be used regardless of origin. He specifically praised models from Chinese firms like DeepSeek, Alibaba, and Moonshot AI, the developer of Kimi K3. The CEO rejected the prevalent security concern that these models could serve as a backdoor for Chinese intelligence, calling it a misconception. He emphasized that companies can download the models, fine-tune them, and apply their own security guardrails within controlled environments.

The context for Huang's comments is the recent launch of Moonshot AI's Kimi K3, an open-weight model released on July 16, 2026. The model has demonstrated...