Anthropic said Tuesday it will donate another $20 million to Public First Action, doubling the Anthropic-backed group’s funding to $40 million since the AI company’s first contribution in February.
Public First Action describes itself as a nonpartisan organization that works with Republicans, Democrats and independents to build public understanding of AI and to advocate for what it calls “sensible AI safeguards.” Anthropic said both donations are earmarked solely for the group’s public-education and policy work and cannot be used to support or oppose any candidate for federal, state or local office.
Tied to Mythos vulnerability findings
Anthropic linked the expanded funding to results from Claude Mythos Preview, the specialized model behind its Project Glasswing security initiative. The company said Mythos Preview has autonomously found thousands of high-severity software vulnerabilities “in every major operating system and browser,” and argued that the same capability shows why increasingly powerful models need government oversight before they are deployed more broadly.
What Anthropic is asking governments to do
The donation announcement points back to Anthropic’s Advanced AI Framework, which calls on governments to independently verify AI developers’ safety claims, impose civil penalties for unsafe practices, and hold authority to slow or block the deployment of models that pose “a serious risk of catastrophic harm.” The framework also calls for export controls on advanced AI chips and tighter rules against illicit model access and distillation attacks, the technique Anthropic has separately accused Alibaba of using to copy Claude’s capabilities.
Part of a broader spending push
The gift adds to a year of rising political spending across the AI industry. Anthropic’s federal lobbying expenditures hit a company record in the second quarter, growing alongside its push for a national AI regulation framework and its ongoing talks with the White House over how frontier models should be overseen. When it made its first donation in February, Anthropic said AI “is being adopted faster than any technology in history, and the window to get policy right is closing.”