Anthropic CEO Dario Amodei goes out of his way to assure President Trump that Anthropic is aligned on AI policy, even as the company faces criticism from the Trump camp.
In a lengthy statement, Amodei highlights Anthropic's close ties with the US government, including a $200 million contract with the Department of Defense to develop advanced AI for national security. The company's language model, Claude, already runs in classified settings through partners like Palantir and the Lawrence Livermore National Laboratory.
Amodei makes a point of publicly supporting Trump's "AI Action Plan." He recalls meeting President Trump at an AI and energy summit in Pennsylvania, "where he and I had a good conversation about US leadership in AI," and notes that other Anthropic executives have participated in White House events focused on AI in healthcare and education.
Throughout his statement, Amodei stresses that Anthropic’s approach is bipartisan and that the company hires experts from both major parties. He says decisions are based on expertise and integrity, not political affiliation.
He also draws a hard line against China, closely following Trump's policy, and emphasizes that Anthropic does not do business with Chinese state-linked firms. The company’s mission, Amodei says, is to help secure US leadership in AI by working with both the government and the opposition.
Anthropic responds to criticism from the Trump camp
Amodei's statement is a direct response to accusations from David Sacks, an AI advisor to President Trump, who claims Anthropic is guilty of "regulatory capture" and is undermining startups with exaggerated warnings.
The criticism focuses on Anthropic's public support for California's Senate Bill 53, which will require more transparency and whistleblower protections for AI companies starting in 2026. Anthropic is the only major AI company to openly back the bill. Reports suggest OpenAI has tried to pressure supporters of the law with legal threats.
Amodei describes supporting SB53 as a practical move, given the lack of progress on AI regulation at the federal level. He says the goal is a clear national set of rules, and the bill is crafted to only affect companies making over $500 million in annual revenue, leaving startups out.
Anthropic co-founder Jack Clark had already called Sacks's accusation "perplexing," saying Anthropic agrees with the government on many issues but has made well-founded counter-proposals on specific points.
