Anthropic lifts funding for AI regulation group to $40mn

Anthropic has committed another $20mn to Public First Action, raising its total support for the AI regulation advocacy group to $40mn, the company said on Tuesday. The donation lands as US officials, states and industry groups contest how far rules for advanced artificial intelligence should go ahead of the fall midterm elections.

The San Francisco-based AI developer said the money is intended for Public First Action’s public education and policy work. Anthropic said both donations were made exclusively for that purpose and cannot be used to influence the election of any federal, state or local candidate.

Anthropic framed the funding as part of its broader push for oversight of frontier AI systems, the term commonly used for the most capable models. In its announcement, the company said it has supported recently enacted state laws requiring more transparency from AI developers, while arguing that disclosure by itself is not enough as leading models become more capable.

Public First Action describes itself as a bipartisan organization focused on public education about AI. The group says it recognizes potential benefits from the technology while advocating safety standards, transparency, accountability measures and protections for whistleblowers who raise concerns about AI risks to policymakers and the public.

Regulatory fight widens

The Wall Street Journal reported on Wednesday that Public First Action has cast itself as a counterweight to Leading the Future, another AI policy group. According to the Journal, Leading the Future’s approach is closer to the White House position and favors rules that are more accommodating to industry development.

The new funding could intensify tensions between Anthropic and the White House, the Journal reported. The company has already faced government pressure: two of its models were briefly shut down over security concerns, the military labeled Anthropic a security risk, and the startup sued the US government over that designation, according to earlier reporting cited by PYMNTS.

The dispute highlights a core policy divide in AI: whether governments should prioritise faster deployment and lower compliance barriers, or impose stronger transparency and safety obligations before the technology becomes more widely embedded in commerce, public services and national security systems.

Critics of Anthropic have argued that the company’s regulatory stance could slow rivals and protect its position, the Journal reported. They have also challenged Chief Executive Dario Amodei’s warnings about AI risks while Anthropic continues to improve its own models. Amodei and Anthropic say their safety concerns are genuine, according to the Journal.

States press ahead

The federal debate is developing alongside state-level enforcement. PYMNTS reported this week that state attorneys general are applying existing consumer protection, privacy and professional licensing laws to AI uses rather than waiting for comprehensive new federal legislation.

That approach means companies deploying AI systems may face scrutiny under current rules covering privacy, discrimination, advertising, consumer protection, professional services and high-impact decision-making, according to the PYMNTS report. For businesses, the practical effect is that AI compliance obligations can arise even before Congress or state legislatures create dedicated AI statutes.

This story draws on original reporting from PYMNTS.