Nvidia CEO, Jensen Huang says Wall Street is "misunderstanding Kimi again," argues American companies should absolutely be allowed to use Chinese AI models.
TL;DR:
Nvidia CEO Jensen Huang argues Chinese open-source AI models like Moonshot AI's new Kimi K3 are excellent and should be freely used — directly opposing Trump administration officials and U.S. AI labs (including Nvidia's own customers OpenAI and Anthropic) who are lobbying to restrict them.
He's confident cheap, open Chinese AI won't push U.S. companies out of the market — he thinks it'll actually expand demand for Nvidia's chips and data centers rather than undercut them.
Context: Kimi K3's release sparked a chip-stock sell-off and comparisons to DeepSeek's January 2025 shock, thanks to its strong performance, low cost, and open weights.
Huang rejects "backdoor" security fears about Chinese models, arguing openness — not restriction — makes AI safer, since outside researchers can spot and patch weaknesses.
He specifically calls out Anthropic, urging it to open up its restricted "Claude Mythos" model rather than keep it locked down.
He's skeptical of the "AI race" framing altogether, expecting the U.S. and China to both be developing AI for the long haul.
He also pushes back on Treasury Secretary Bessent's talk of sanctions over alleged Chinese IP theft, saying learning/distillation from other models is a normal part of AI development — enforcement should target misconduct, not the technology itself.
Full Article: https://www.axios.com/2026/07/22/nvidia-jensen-huang-china-open-source-ai