Anthropic is donating another $20 million to Public First Action

Details

Anthropic has contributed an additional $20 million to Public First Action, a nonpartisan organization focused on AI education and policy, bringing its total support to $40 million. The donation cannot be used to influence candidate elections and targets public education and policy missions specifically. In explaining the timing, Anthropic cites Claude Mythos Preview's earlier discovery of thousands of high-severity vulnerabilities in every major operating system and browser — released only to trusted cyber defenders through Project Glasswing. The company uses this as evidence that frontier model capabilities now require governments to begin building regulatory capacity before deployment, not after. The donation aligns with Anthropic's Advanced AI Framework, which calls for government ability to verify company safety claims, enforce practices through civil penalties, and block AI deployments posing catastrophic risk. Anthropic has supported transparency laws in several states.