Jensen Huang says US shouldn't fear China's open-source AI, openness is safer

Nvidia CEO Jensen Huang said in a recent interview that the US shouldn't fear China's open-source AI models, but should instead worry about domestic calls to ban them. He believes great open-source models should be used, and US companies should be allowed to use Chinese open-source AI. Huang disagrees with OpenAI and Anthropic's concerns about Chinese open-source models, saying open-source models can expand the market, and users will still pay for reliable closed-source services. He stressed there's zero chance Chinese companies will push US firms out of the market. Regarding the recent market shock from Moonshot AI's Kimi K3 open-source model, Huang thinks the market misunderstood its impact. Cheaper, more open AI models will increase demand for chips and data centers, benefiting infrastructure providers. He also argued that restricting open-source models in the name of national security makes the US more vulnerable, and openness actually means higher security.