Anthropic faces Silicon Valley backlash for refusing to sign open-weight AI letter

The tech industry is rallying around open-weight AI, but one major lab is staying conspicuously silent.

Anthropic is drawing criticism from across Silicon Valley for not signing an open letter advocating for open-weight AI, as Washington considers restrictions on some Chinese models.

The Open Letter

On Friday, over two dozen companies — including Nvidia, Meta, and Microsoft — signed "Open Weights and American AI Leadership," endorsing open-weights, which means developers can download, modify, and deploy AI models on their own infrastructure.

Over the weekend, several other major names in AI signed the letter: OpenAI, Google, and SpaceX.

As of Monday, the only major frontier AI lab not to back the industry-wide defense of open models is Anthropic — and the Claude maker is facing significant criticism for it.

Industry Reactions

"The entire tech industry (save for Anthropic) has come out in favor of open source AI," said David Sacks, a venture capitalist and former AI czar for the Trump administration.

"They won't stop until they kneecap open source. The rest of the industry needs to watch these guys like a hawk," he wrote in a post on X.

"Proud that @OpenAI signed the letter. Ant is silent on it," wrote Peter Steinberger, an OpenAI employee and the founder of agentic platform OpenClaw, in his own X post.

Bill Gurley, a general partner at VC firm Benchmark, also criticized Anthropic for not signing the letter and suggested the company was attempting to preserve its position in the AI race.

"Excited to see Anthropic acknowledge the real problem with open is it competes with their corporate economic strategy. The 'awareness' here is revealing," wrote Gurley in an X post responding to an Anthropic employee.

"Who DIDN'T sign the Open Weight letter is far more interesting than who did," wrote Kai-Fu Lee, a former Google and Microsoft exec who founded Chinese open-model startup 01.AI, while tagging Anthropic's Claude AI in his X post.

Anthropic's Position

Anthropic did not immediately respond to a request for comment.

Anthropic CEO Dario Amodei has previously warned that developers lose control of frontier AI models once their weights are released. Anthropic has instead favored gated access to powerful models, launching Project Glasswing, an initiative to give vetted companies early access to its powerful cyber models to fix security flaws, earlier this year.

China Dialing Up the Open Model Debate

The debate over open models has been supercharged by the recent launch of Kimi K3, an open-weight model built by the Chinese AI lab Moonshot that rivals the performance of top US closed models on some benchmarks.

Last week, a top White House official accused Moonshot of using Anthropic's Fable AI model to train K3, a practice known as distillation, while Treasury Secretary Scott Bessent suggested the US government could sanction Chinese open-source AI models that are found to have distilled their US rivals. China has, broadly, pursued a more open approach to AI than the US.

The argument for open-weight AI was boosted last week after Hugging Face said it used a model from China's Z.ai in its defense against a hack from OpenAI's rogue AI agent. Hugging Face said it was unable to use closed AI models because of restrictions on them.

The outpouring of support for open models continued on Monday with Nvidia announcing the launch of the Open Secure AI Alliance, an initiative advocating for open AI models in cybersecurity, which was backed by Microsoft, SpaceX, and other tech firms.