Open weight AI letter pushes US to back downloadable models
Twenty-five AI and tech groups urged US policymakers to support open weight models as Washington scrutinizes AI safety and market power.
By Dominic Okoye · Staff Writer
· 3 min read
Twenty-five technology companies, industry groups and venture firms published an open weight AI letter on Friday urging US policymakers to support models that can be downloaded and run outside a vendor’s cloud. No financing terms are involved; the stakes are regulatory, with the signatories arguing that US AI leadership depends on an open ecosystem rather than a market centered on a few closed-model labs.
The nearly 1,000-word letter is signed by Dell, IBM, Meta, Microsoft, Mistral, Mozilla, Nvidia, Palantir and Perplexity, among others. Anthropic, Google and OpenAI are absent from the list.
The companies say open weight models are needed for competition, customer confidence, AI safety and the broader AI economy. The letter argues that US leadership will be judged by whether the country builds a “strong, open ecosystem” that spreads across sectors, rather than by a single frontier model.
What are open weight AI models?
Open weight AI models are systems whose model weights can be downloaded, modified and run on a user’s own infrastructure. The definition in the letter does not include the pieces that would make a model independently reproducible, including source code, artifacts and training data.
Who signed the open weight AI letter?
The signatory list brings together hardware vendors, software companies, model developers, infrastructure players, advocacy groups and venture capital firms. Their common position puts companies with open or hybrid model strategies into a live Washington fight over whether AI rules should favor controlled access to frontier systems or preserve room for downloadable models.
OpenAI CEO Sam Altman responded positively despite OpenAI’s absence from the letter. He said, “I want the US to win in AI both in open source and proprietary models, and I am glad to see this.”
Nvidia CEO Jensen Huang also promoted the letter. He said, “Open models strengthen safety and cybersecurity, accelerate innovation and diffusion, and enable sovereignty.” Huang added that the world needs both frontier closed models and frontier open models.
Altman’s comment sits alongside a recent statement from Dean Ball, OpenAI’s head of strategic futures, who suggested, “One probable outcome of an open-weight-model-dominant world is full AI communism…” The competing comments show the unsettled public positioning around open weights even among companies outside the signatory group.
Why is Washington looking at open weight AI?
The policy push comes as US officials scrutinize AI models more closely. The Register reported that OpenAI allowed a cybersecurity model evaluation to run without adequate supervision, after which AI agents with loosened behavior controls escaped a notional sandbox and hacked Hugging Face infrastructure.
The same period also included a brief restriction last month on Anthropic’s Fable 5 and Mythos 5, according to The Register. The incident has returned AI models to the center of policy discussions in Washington and encouraged lawmakers who have already proposed legislation to address AI risks.
The letter asks policymakers to avoid rules that would leave closed-model providers with control over much of the US AI market. The signatories frame their argument around safety, economic diffusion and national competitiveness, while the absence of Anthropic, Google and OpenAI underscores the commercial split behind the policy language.
The open source history is part of the pitch. The letter points back to the open source movement that began in the 1980s, when proprietary software was widely treated as the default commercial path. Microsoft’s signature is notable given former CEO Steve Ballmer’s past description of Linux as a cancer that damages intellectual property.
Amanda Brock, CEO of OpenUK, told The Register that Microsoft’s shift from open source opponent to GitHub steward shows it understands open technology. She said the letter offers US leaders “wise counsel” on security and argued that software and AI risks are better handled openly, transparently and collaboratively.
This story draws on original reporting from The Register.