Nvidia CEO backs open-weight AI models against OpenAI, Anthropic

Nvidia CEO Jensen Huang has published an open letter backing open-weight artificial-intelligence models—AI systems where the code and trained weights are made publicly available—as viable competitors to the proprietary systems developed by OpenAI and Anthropic. The letter represents a public statement of support from one of the industry's most influential voices; Nvidia supplies the chips that power the vast majority of large language model training and inference worldwide. By endorsing open-weight models publicly, Huang is signaling that Nvidia views this category as strategically important and worthy of support, even as it challenges the closed, commercial models that have dominated recent AI headlines. The article does not detail the specific contents of the letter, name particular models or companies Huang endorsed, or indicate when it was published, leaving readers without the precise claims or scope of his support.

Nvidia's dominance in AI chip manufacturing gives it outsized influence over which AI systems succeed commercially. By issuing a public letter in support of open-weight models—AI systems whose underlying code and model weights are publicly released, allowing independent researchers and companies to use, study, and modify them—Huang is taking a visible stance in a growing divide within AI development. OpenAI and Anthropic build proprietary, closed systems where the model internals remain secret; open-weight alternatives (including those from Chinese developers, which the headline references) offer transparency and independence from any single company's control. Huang's endorsement suggests Nvidia sees commercial opportunity and strategic value in supporting this ecosystem, even as it competes with closed models that also depend on Nvidia hardware.

No comments yet. Be the first to share your thoughts!