A coalition of American technology companies is publicly urging Washington not to clamp down on "open-weight" AI models — systems whose underlying parameters anyone can download, run and build on.
According to Reuters, Nvidia, Microsoft and other tech heavyweights made their case to lawmakers on Friday, wading into a growing policy debate. Tom's Hardware reports that 25 companies in total signed a joint letter — Nvidia plus 24 others — spanning chipmakers, server vendors, cloud operators, enterprise software firms, security companies and venture funds. Signatories named across the coverage include Meta, Dell, IBM, Hugging Face and the Linux Foundation.
CNBC reports the group warned against "premature restrictions" on open-weight models. The Times of India, citing the letter, quotes the argument that "openness may be one of the most important paths to AI safety." Backers say keeping models open is crucial to maintaining US leadership in AI.
The timing is pointed. Tom's Hardware notes the letter lands as Washington weighs a ban on Chinese AI models, and Bloomberg ties the push to the recent launch of Kimi, a Chinese open model. Nvidia warns, per Interesting Engineering, that US curbs could backfire by pushing AI innovation toward China. Nvidia CEO Jensen Huang shared the letter in his first-ever post on X.
Just as notable is who did not sign. Tom's Hardware and Benzinga highlight that OpenAI, Anthropic, Google and Amazon — several of them associated with more closed, proprietary models — are absent from the list. Even so, Seeking Alpha reports that OpenAI's Sam Altman voiced support for Huang's stance on the importance of open models in the AI race.
Why it matters: the fight over open-weight AI will help decide whether cutting-edge models stay freely available to researchers and startups — or become tightly controlled — shaping both US competitiveness against China and who gets to build the next wave of AI.