Anthropic is contributing an additional $20 million to Public First Action, bringing its total support for the organization to $40 million. This donation supports the non-partisan group's public education and policy mission regarding AI safeguards.
- The funds are restricted to public education and policy work and cannot be used to influence elections.
- Anthropic cites rapid AI improvements, including Claude Mythos Preview discovering thousands of high-severity vulnerabilities, as a key driver for urgent policy action.
- The company advocates for the Advanced AI Framework, which includes government verification of safety claims, civil penalties, and independent model evaluation.
- Anthropic also supports tightening export controls on advanced chips to maintain American AI leadership against authoritarian competitors.
Anthropic views this donation as a way to raise the salience of urgent policy debates needed to mitigate risks while capturing the benefits of frontier AI.