Anthropic is significantly expanding its political advocacy campaign ahead of the November U.S. midterm elections, committing an additional $20 million to Public First Action, a nonprofit organization focused on artificial intelligence policy, bringing its total contribution to $40 million.
The latest funding underlines the intensifying battle among leading AI companies to shape the regulatory framework that will govern advanced AI systems in the United States. As frontier AI models become increasingly capable and Washington prepares for a new round of legislative debates after the midterm elections, technology firms are investing heavily in organizations that can influence the national conversation around AI safety, innovation and competitiveness.
Anthropic said the donation is intended to elevate the urgency of AI governance rather than support individual political candidates.
Register for Tekedia Mini-MBA edition 20 (June 8 – Sept 5, 2026).
Register for Tekedia AI in Business Masterclass.
Join Tekedia Capital Syndicate and co-invest in great global startups.
“Our donation to Public First Action is one way in which we’re trying to raise the salience of this urgent policy debate,” the company said in a statement.
Anthropic argued that the rapid evolution of powerful AI models, including its latest Claude Mythos system, has heightened the need for policymakers to establish safeguards before the technology becomes even more deeply embedded across the economy.
“We need policymakers and candidates to put forward measures that mitigate risks,” the company added.
The investment comes as artificial intelligence has become one of the defining policy issues in Washington, joining national security, semiconductor leadership and energy infrastructure as areas viewed as critical to maintaining U.S. technological dominance. Lawmakers are weighing how to regulate frontier AI without undermining American competitiveness, particularly as China accelerates investment in advanced AI capabilities.
Unlike traditional political action committees, Public First Action is organized as a nonprofit and is not legally required to disclose its donors. The organization announced in June that it had raised $80 million, positioning it among the most well-funded AI-focused advocacy groups operating in Washington.
Anthropic said its contribution “cannot be used to influence the election of any candidate for federal, state, or local office.” However, nonprofits operating under this structure can finance issue-based advocacy campaigns that highlight elected officials and candidates based on their policy positions. Such campaigns can shape public perception without explicitly urging voters to support or oppose a candidate.
Public First Action’s first advertising campaign highlighted Republican Senator Marsha Blackburn’s work on artificial intelligence and online child safety legislation. Blackburn is leaving the Senate to pursue the Tennessee governorship.
The organization was co-founded by former Republican Congressman Chris Stewart of Utah and former Democratic Congressman Brad Carson of Oklahoma. The pair also established allied super PACs, Defending Our Values PAC and Jobs and Democracy PAC, which support Republican and Democratic candidates, respectively.
Public First Action has emerged as one side of a visible divide within the AI industry over how aggressively governments should regulate advanced AI. The group has positioned itself against Leading the Future, a rival super PAC backed by influential technology figures including OpenAI President Greg Brockman and venture capitalists Marc Andreessen and Ben Horowitz. OpenAI has said Brockman is supporting the organization in his personal capacity rather than on behalf of the company.
Leading the Future has raised more than $75 million, highlighting the growing flow of money from Silicon Valley into AI-related political advocacy. The emergence of competing advocacy groups indicates that the industry’s policy debate extends beyond Congress to competing visions among AI developers themselves over how the technology should be governed.
Among major AI executives, Anthropic Chief Executive Dario Amodei has become one of the most vocal advocates of stringent federal oversight for frontier AI systems. He has repeatedly warned that the capabilities of next-generation AI models could outpace existing regulatory institutions if safeguards are not implemented before deployment.
In an essay published last month, Amodei proposed creating a federal regulator modeled after the Federal Aviation Administration that would oversee advanced AI development. Under his proposal, developers of the most powerful AI models would be required to undergo mandatory safety testing before releasing new systems, with regulators empowered to delay or block deployment if evaluations identified unacceptable risks.
His proposals go beyond the voluntary safety commitments currently favored by many technology companies. They also contrast with the approach advocated by President Donald Trump’s administration and several AI industry leaders, who have warned that mandatory pre-release government reviews could hamper innovation, slow commercialization and weaken the United States’ competitive position against geopolitical rivals, particularly China.
The policy disagreement reveals a broader divide within the AI sector. Anthropic has consistently argued that the risks posed by increasingly autonomous AI systems warrant stronger government oversight, while several competitors have emphasized flexible, innovation-friendly regulation that relies more heavily on industry standards than binding federal mandates.
Amodei has also personally backed the advocacy effort. In May, he donated $1 million to Public First, the bipartisan super PAC affiliated with Public First Action.