Anthropic Gives $40 Million to Super PAC as Congress Weighs AI Rules

Anthropic has contributed $40 million through two donations to the super PAC Public First Action, adding major political funding to its campaign for federal AI safeguards. The money campaign lands as lawmakers debate whether guardrails for increasingly capable AI systems would improve safety or favor the companies best able to meet them.

One reported $20 million payment arrived in late July. Anthropic has said it supports independent AI evaluations, transparency requirements and the authority to block risky deployments. Those are concrete tools for oversight, but they also imply compliance obligations that could matter differently for large and small developers.

A broader political spending push

The super PAC contribution is part of a wider reported effort. Anthropic’s July lobbying disclosure listed $2 million in spending over three months—$400,000 more than its April report and $1 million more than its spending in the final quarter of 2025. Its political action committee also gave $350,000 to Republican and Democratic lawmakers, including $15,000 to the National Republican Congressional Committee.

$40 millionSuper PAC contributions

Anthropic contributed $40 million through two donations to Public First Action, including a reported $20 million payment in late July.

$2 millionJuly lobbying disclosure

The reported amount was $400,000 above Anthropic’s April disclosure and $1 million above its final-quarter 2025 spending.

$350,000PAC donations to lawmakers

Anthropic’s PAC distributed the money across both parties, including $15,000 to the National Republican Congressional Committee.

Public First Action has already put some of its resources into electoral politics. It gave $3 million to a PAC supporting New York Assemblyman Alex Bores, part of a reported $10 million effort by Anthropic-affiliated groups. Bores later lost to fellow Democrat Micah Lasher.

Safety policy meets an incumbent question

The policy divide

• Anthropic is advocating outside evaluation, transparency rules and authority to stop risky deployments.
• Sens. Ron Johnson and Mike Lee have questioned whether AI rules supported by leading companies could hinder competition and benefit established firms.
• Rep. Ted Lieu has called for committee action on legislation involving AI guardrails and a proposed kill switch.

That disagreement is the central test for any congressional response. A rule can demand more evidence before a risky system is deployed, yet still impose costs that are easier for a well-funded company to absorb. Johnson and Lee’s criticism does not establish that Anthropic’s preferred rules would have that outcome; it identifies the competitive risk lawmakers will need to weigh alongside the company’s safety case.

Congress has not chosen a path

Congress now faces pressure to translate broad safety concerns into specific requirements. Lieu’s call for committee action points toward a more active legislative response, while Anthropic’s political spending means the company will be a visible participant in deciding what that response looks like. The unresolved question is whether lawmakers can build rules that address risky deployments without turning compliance capacity into a barrier to entry.