AI startup Anthropic has doubled its midterm election-related spending to $40 million, intensifying the clash within Silicon Valley between opposing camps on how to regulate the transformative technology.

Anthropic announced on Wednesday an additional $20 million investment into the Public First Action group. This political organization advocates for government-mandated safety standards for high-performance AI models and requires developers to publicly disclose their technologies' risks. Earlier this year, Anthropic had already contributed a similar amount to the Public First organization, prior to the launch of its Mythos model, as debates over AI's military applications spurred widespread calls for increased oversight.

Public First Action stated its goal is to counterbalance another major AI funder, "Lead the Future," which is backed by several tech giants, including an executive from OpenAI and venture capital firm Andreessen Horowitz. "Lead the Future" holds a regulatory stance closer to the White House, favoring industry-friendly rules that impose fewer restrictions on AI development.

Anthropic and its CEO, Dario Amodei, have consistently opposed this approach, warning that greater government intervention is necessary to prevent catastrophic risks from artificial intelligence. Amodei donated $1 million in May to a super PAC funded by Public First Action. Other employees at Anthropic have also continued to contribute funds to causes supporting AI regulation.

The company stated on Wednesday, "We need policymakers and candidates to implement risk-reducing safeguards."

This new round of funding is likely to further escalate tensions between Anthropic and the White House. The discord has deep roots: two of Anthropic's models were previously shut down over safety concerns, and the U.S. Department of Defense has listed the company as a national security risk, a designation Anthropic is currently challenging in court.