AI Ethics & Regulation

Microsoft, Nvidia and Meta Defend Open-Weight AI—Anthropic and OpenAI Refuse to Sign

After Washington accused China's Moonshot AI of IP theft, 25 organizations defended open weights. Anthropic and OpenAI declined to sign as the industry splits over Chinese models.

The fight over open-weight AI entered a new and unusually public round this week, and it split the technology industry along a revealing line: open versus closed. On July 22, White House officials accused China’s Moonshot AI of stealing American intellectual property through distillation and floated sanctions, with Anthropic’s policy chief calling the alleged operation “industrial espionage.”

The pushback was swift. Nearly 200 companies, including Y Combinator and Proton, co-signed a letter to President Trump through the new Little Tech Association urging him not to cut off the open Chinese models they rely on, asking for “a scalpel rather than a sledgehammer.” On Friday, a statement signed by 25 major organizations warned Washington against “premature restrictions” on open weights.

The signatories of that statement, titled “Open Weights and American AI Leadership,” include Microsoft, Nvidia, Meta, Hugging Face, IBM, Mozilla, Palantir, Perplexity, Mistral, and Y Combinator. Two names are conspicuously absent: Anthropic and OpenAI did not sign, according to The New Stack.

Why developers already chose open-weights AI models

For startups, the stakes are not abstract — America’s frontier labs have priced many of them onto Chinese open weights. “There’ll be hundreds of companies that instantly die,” Particle founder Shuail Doshi told The New Stack of a potential ban. “It’s great for Anthropic. We’re all going to have to spend money on Anthropic.”

The math explains the panic. Moonshot’s Kimi K3 is a near-frontier model, and a New Stack head-to-head found it wrote the same code as Anthropic’s Claude Fable 5 across a bug fix, a refactor, and a feature build for about a third of the cost — though it ran roughly four times slower. For a startup counting tokens, that is often an easy trade.

Developers have already voted with their tokens. On OpenRouter, Chinese models have accounted for more than 30 percent of token usage by US customers every week since February, peaking at 46 percent. OpenRouter is only one slice of the market, but it captures the direction of travel toward much cheaper Chinese models.

An industry talking its book

Anthropic has taken the clearest position, and Axios reports OpenAI has aligned with it in warning Washington about powerful Chinese open-weight models. The White House, meanwhile, is trying to draw a narrower line — defending open weights generally while punishing Chinese companies it believes used distillation to steal American capabilities.

Across from them are the companies that supply the chips, clouds, tools, and applications the AI economy runs on. Their statement calls distillation “a widely used technique” and warns policymakers not to conflate legitimate model development with unlawful extraction. Nvidia CEO Jensen Huang, asked by Axios whether open labs should be allowed to distill closed models, answered that distillation is “fundamental to intelligence.”

Everyone here is arguing their own interests. Nvidia sells chips to Anthropic and OpenAI as eagerly as to anyone running open models, and Microsoft is one of OpenAI’s largest investors — these are not companies with a simple grudge against the closed labs. No ban has been drafted; the White House calls the reporting “baseless speculation,” and the distillation allegation has not been backed by publicly verifiable evidence. But the sides are now on the record.

FAQ

Which companies defended open-weight AI models?

A statement titled “Open Weights and American AI Leadership” was signed by 25 organizations including Microsoft, Nvidia, Meta, Hugging Face, IBM, Mozilla, Palantir, Perplexity, Mistral and Y Combinator, warning against premature restrictions.

Why didn’t Anthropic and OpenAI sign?

Both have taken the opposite position. Anthropic has been the clearest voice warning Washington about powerful Chinese open-weight models, and Axios reports OpenAI has aligned with it. Neither signed the open-weights defense statement.

What triggered the open-weights fight?

On July 22, White House officials accused China’s Moonshot AI of stealing US intellectual property through distillation and floated sanctions, prompting nearly 200 startups and 25 major organizations to push back within 48 hours.