Over 60 US tech giants urge open-weight AI models; Anthropic CEO warns of cybersecurity risks

Published Jul 30, 06:10 AM UTC

Over 60 US 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 US leadership and counter Chinese competition, specifically referencing Moonshot's recently launched Kimi K3 model. The letter argues open-weight models enhance accessibility, adaptability, cybersecurity defense, and transparency. It calls on policymakers to expand compute access for startups, invest in shared training assets, and avoid premature restrictions. Notably, Anthropic and Amazon did not sign. On July 27, Anthropic CEO Dario Amodei published a blog post clarifying the company does not support banning open-weight models but warned of risks from hostile states using such models for cyber-attacks or biological weapons. The debate highlights tensions between open innovation and national security concerns in the AI sector.