Over 60 US tech giants sign open letter backing open-weight AI models; Anthropic CEO warns of security risks

Published Jul 31, 07:56 AM UTC

Over 60 U.S. tech companies, including Microsoft, Nvidia, OpenAI, SpaceX, Google, IBM, Palantir, and Meta, signed an open letter on July 24, 2026, advocating for American open-weight AI models to maintain U.S. leadership and counter Chinese advances like Moonshot's Kimi K3. The letter argues open-weight models enhance accessibility, adaptability, and cybersecurity through transparency and vulnerability detection. It urges policymakers to expand compute access for startups and avoid premature restrictions. Anthropic, notably absent from the signatories, had its CEO Dario Amodei publish a blog post on July 27 clarifying the company does not support a ban but warning that open-weight models could be exploited by hostile states for cyber or biological attacks. The debate highlights tensions between open innovation and security concerns in the AI industry.