Key facts
- Anthropic is facing criticism for not signing an open letter supporting open-weight AI models.
- Major AI companies including Nvidia, Meta, Microsoft, OpenAI, and Google have endorsed the open-weight initiative.
- Anthropic CEO Dario Amodei has previously voiced concerns about the release of frontier AI model weights.
- The debate over open models has been amplified by the launch of China's Kimi K3 model.
- Nvidia launched the Open Secure AI Alliance to advocate for open AI models in cybersecurity.
Anthropic is facing significant criticism from the tech industry for its decision not to sign an open letter supporting open-weight artificial intelligence models. The initiative, titled "Open Weights and American AI Leadership," has garnered endorsements from over two dozen companies, including major players like Nvidia, Meta, Microsoft, OpenAI, Google, and SpaceX.
The absence of Anthropic's signature makes it the sole major frontier AI lab not to publicly back the defense of open models. This stance has drawn sharp reactions from venture capitalists and industry figures. David Sacks, a venture capitalist, accused Anthropic of attempting to undermine open-source AI, urging the industry to monitor the company closely. Peter Steinberger, an OpenAI employee, highlighted Anthropic's silence on the matter, while Benchmark partner Bill Gurley suggested the company's position is driven by its corporate economic strategy.
Anthropic CEO Dario Amodei has previously expressed concerns regarding the potential loss of control over advanced AI models once their weights are released. The company has instead focused on providing gated access to its powerful models, including initiatives like Project Glasswing for cybersecurity assessments.
The debate over open versus closed AI models has been intensified by recent developments, such as the launch of Kimi K3, an open-weight model from Chinese AI lab Moonshot. A U.S. White House official alleged that Moonshot used Anthropic's Fable AI model for training K3, a practice known as distillation. Treasury Secretary Scott Bessent indicated that the U.S. government might consider sanctions against Chinese open-source AI models found to have distilled U.S. rivals.
In a related development, Nvidia announced the launch of the Open Secure AI Alliance, an initiative advocating for open AI models in cybersecurity, which has received backing from Microsoft, SpaceX, and other tech firms.