Nvidia's push to rally the tech industry behind "open weights" artificial intelligence has picked up momentum. According to Forbes, the number of companies signing Nvidia's open weights letter has doubled to 50 — a notable show of support for the idea that AI models should be released more openly.
But Forbes also flags two conspicuous absences: Amazon and Anthropic are not among the signatories. Both are major players in the AI race, which makes their decision to stay off the list one of the more telling details of the story.
A quick primer for non-engineers: "open weights" refers to AI models whose trained internal parameters — the numerical values a model learns during training — are made publicly available, rather than locked behind a company's servers. Backers argue this openness fosters competition, transparency, and broader access. Critics worry about safety and misuse. Where a company lands on that spectrum can signal how it wants the AI ecosystem to be structured.
Nvidia, whose chips power much of the AI industry, benefits when more developers build and run models — so its championing of open approaches aligns with its commercial interests as a hardware supplier.
The growth from a smaller group to 50 names suggests Nvidia's framing is gaining traction across the sector. Yet the gaps matter as much as the additions: an industry "consensus" that excludes heavyweights like Amazon and Anthropic is not really a consensus.
Why it matters: which companies line up behind open weights — and which pointedly don't — offers an early map of how the AI industry may split over openness, control, and who gets to shape the technology's future.