Nvidia CEO Jensen Huang says U.S. firms should be allowed to use Chinese AI models

Huang said Washington should not restrict open-source AI systems from Chinese developers such as Kimi, arguing strong open and closed models can coexist and broaden AI adoption.

36d ago

Fact Check

The claim is confirmed by the primary Axios exclusive interview and corroborated by Fortune and Yahoo Finance, all dated July 22, 2026. Huang stated U.S. firms should be allowed to use Chinese open-source models like Kimi (Axios: 'excellent open-source models should be used'; Yahoo: companies should 'absolutely' be allowed), that Washington should not restrict/ban them, and that open and closed models coexist and broaden AI adoption (Fortune: industry needs both open and closed models). Every element of the claim matches the sourced reporting.

Summary

Nvidia CEO Jensen Huang said U.S. companies should be free to use Chinese open-source AI models and urged Washington not to ban or restrict them. Speaking to Axios on July 21, Huang said Chinese models such as Kimi should remain available to U.S. businesses and argued that strong open-source systems can coexist with closed models. He said wider access to capable AI would expand usage and in turn increase demand for computing, data centers and services. The comments add to a broader debate in Washington over whether Chinese AI models should face limits in the U.S. market amid competition and security concerns.

Terms & Concepts

open-source AI models: AI systems whose underlying model can be used and adapted more openly by developers or companies.
closed models: AI systems controlled by their creators, with access and modification typically limited by the provider.