Jensen Huang calls for US-China cooperation on AI safety

By Splitfeed EditorialLast updated Oct 16, 2026, 07:00 UTC

On October 15, 2026, Nvidia CEO Jensen Huang stated publicly that the United States must cooperate with China on artificial intelligence safety, describing AI as a technology that is potentially dangerous. Speaking in remarks that also included a call for UK participation in that dialogue, Huang went further by saying Nvidia itself would halt chip supply to any AI developers it considers unsafe. He also broke with OpenAI CEO Sam Altman over cybersecurity, rejecting Altman's position that the industry should accept certain AI cybersecurity risks and insisting instead that the industry should do everything possible to address them.

The significance of Huang's position derives from where Nvidia sits in the AI supply chain. Its chips are foundational to AI development globally, which means a threat to cut supply to unsafe developers is not symbolic posturing but a lever with real consequences. At the same time, his call for US-China cooperation runs directly against the direction of US policy, which has moved steadily toward restricting advanced chip exports to Chinese firms. Whether international AI safety coordination is even politically viable under those conditions is genuinely contested, and his public disagreement with Altman over cybersecurity risk tolerance draws a visible line between two of the industry's most powerful figures on a question that has no settled answer.

With no published reactions available yet from the Pro-AI, Anti-AI, or Middle Ground camps, their positions can only be anticipated based on the nature of the story. Pro-AI voices would typically welcome Huang's framing of safety as a shared global challenge while highlighting his continued commitment to AI development. Anti-AI voices would likely treat his admission that AI is potentially dangerous as a concession that validates their concerns, while questioning whether voluntary supply cuts are an adequate safeguard. Middle Ground observers would probably focus on whether meaningful US-China safety cooperation is realistic given the current geopolitical climate, and whether Huang's proposed mechanisms, such as cutting chip access, could be applied consistently and fairly.

The argument that most needs watching is whether Nvidia moves to operationalize its stated policy of withholding chips from developers it deems unsafe, since the criteria it uses and the first cases it applies them to would clarify how much weight the commitment actually carries. Any formal US or UK government response to Huang's call for trilateral dialogue would also signal whether his proposal finds institutional traction or remains an industry wish.