Anthropic has announced an additional $20 million donation to Public First Action, a non-partisan organization focused on AI policy education, bringing their total support to $40 million. This funding aims to promote sensible AI safeguards and policy measures as AI models rapidly advance.
TL;DR
- Anthropic increases funding for Public First Action to $40 million to support AI policy education and advocacy.
- The donation underscores the urgency of AI policy as models like Claude Mythos Preview uncover critical vulnerabilities.
- Anthropic advocates for transparency, safety testing, and government oversight to mitigate AI risks.
What happened
Anthropic has contributed an additional $20 million to Public First Action, a non-partisan organization dedicated to educating the public about AI and promoting sensible AI safeguards. This brings their total support to the organization to $40 million. Both donations are exclusively earmarked for Public First Action's public education and policy mission, with no funds used to influence any candidate's election.
The initial $20 million donation in February 2026 aimed to support policies maintaining AI safeguards, sustaining American AI leadership, and demanding transparency from AI developers. The recent donation reflects the growing need for these policies as AI models advance rapidly.
Why it matters
The rapid improvement of AI models, such as Claude Mythos Preview, highlights the need for robust policy frameworks. Earlier this year, Claude Mythos Preview identified thousands of high-severity software vulnerabilities, emphasizing the potential risks if such models fall into the wrong hands. This underscores the importance of proactive policy measures to mitigate AI risks.
Anthropic's Advanced AI Framework outlines key policy proposals, including transparency, safety testing, and government oversight. These measures are crucial for ensuring that AI benefits, such as accelerating scientific progress and economic growth, are realized while minimizing risks. The donation to Public First Action aims to raise awareness and drive policy discussions forward.
Key facts
- Anthropic's total support for Public First Action now stands at $40 million.
- Public First Action is a non-partisan organization focused on AI policy education and advocacy.
- The donations are exclusively for public education and policy missions, not for influencing elections.
- Claude Mythos Preview discovered thousands of high-severity software vulnerabilities, including in major operating systems and browsers.
- Anthropic's Advanced AI Framework proposes transparency, safety testing, and government oversight for AI developers.
- The donation aims to support policy efforts to tighten export controls on advanced chips and semiconductor manufacturing equipment.
- Anthropic also supports AI for Science research grants and has committed $10 million to Canadian AI research.
Context
Anthropic's increased funding for Public First Action comes at a critical time as AI models become more capable and the need for robust policy frameworks grows. The organization's efforts to educate the public and policymakers about AI risks and benefits are essential for shaping effective policies.
As AI advancements accelerate, the national security stakes of AI are also rising. Ensuring that democracies maintain their lead in AI technology and prevent authoritarian misuse is a key focus for Anthropic. The donation to Public First Action is part of a broader strategy to promote responsible AI development and policy.
