Anthropic Adds $20 Million to Public First Action Amid AI Safety Push

Anthropic said on Wednesday it will donate an additional $20 million to bipartisan advocacy group Public First Action, bringing its total pledged contribution to $40 million. According to ChainCatcher, the company said the decision was driven by the growing capabilities of AI models, including its in-house Mythos model, and the potential risks they may create. Public First Action is a 501(c)(4) nonprofit focused on public education around AI transparency and safety. The group is linked to three super PACs and plans to support candidates in both parties who back AI safeguards. Anthropic said the donation is intended only to support Public First Action's policy mission and cannot be used to influence any specific candidate's election. Anthropic said AI policy decisions will affect the labor market, national security, and other areas of public life, and it wants to promote flexible regulation that allows AI benefits to be realized while risks are managed. The company also noted that super PAC Leading the Future, which supports rapid AI development and lighter regulation, has received significant funding from backers including OpenAI co-founder Greg Brockman and venture firm a16z. As AI becomes an election issue, public education and lobbying spending by advocacy groups is increasing.