Trump adviser's memo puts Dario Amodei at the root of EA

A memo by a Trump political adviser accuses effective altruism of putting "foreigners over citizens, shrimp over families, future hypothetical people over the living," and places Anthropic CEO Dario Amodei at the movement's foundation, Axios reports. The document began circulating in the White House this week and says the movement "built the AI-doom pipeline."

At a glance

• Trump allies are casting Amodei as the face of AI "doomerism," and Axios sources say they see him as an easy foil because of his politics and his focus on AI safety.
• The memo names Dario and Daniela Amodei among those who "built the [EA] network," labels those ties "The Anthropic knot," and lists abortion, veganism and "digital minds" rights without tying them to him.
• The catch for Anthropic is timing: Axios calls the attacks worrisome for investors ahead of an expected record-setting IPO, and notes that OpenAI catching up could make the company an easier target.

If you missed the spring, Axios describes Anthropic's relationship with the White House as rocky. According to Tech Policy Press, Defense Secretary Pete Hegseth gave Amodei a deadline of 5:01 p.m. on Friday, February 27, to allow unrestricted use of Anthropic's models "for all legal purposes," and Anthropic refused over concerns about autonomous weapons and mass surveillance. The same outlet reports that on February 27 President Trump directed federal agencies to stop using Anthropic's products, and Hegseth designated the firm a supply chain risk.

A Trump adviser's memo calls the Amodei family ties "The Anthropic knot"

According to Axios, a Trump political adviser wrote the memo, and it was drafted with Trump in mind. It paints effective altruism as a fringe, cultish collective out of touch with mainstream America. It names Dario Amodei and his sister Daniela, Anthropic's president, among those who "built the [EA] network." Daniela Amodei's husband once led a philanthropy dedicated to AI safety.

The memo says that on its current agenda, the movement puts "possible machine minds over Americans." It lists abortion, veganism, devaluing human life by overvaluing animals, and civil rights-like protections for "digital minds." According to Axios, the memo does not tie these ideas to Amodei directly. It also mentions Sam Bankman-Fried without linking him to any of them.

In the view of the Florida-heavy White House and its allies, Anthropic has too many Biden-era liberals, Democrats and strange Californians on staff, Axios writes. Months ago, some at Anthropic asked about a meeting between Amodei and the president. White House officials advised against it because "Dario's a little too weird" for Trump, one senior official said.

Daniela Amodei told Wired in 2025 that she does not identify with the EA label

Anthropic has been working to distance itself from the movement. Daniela Amodei told Wired in 2025: "I don't identify with that terminology." Amanda Askell, an early employee and philosopher by training who helps develop Claude's constitution, says EA is "not a theme of the organization." Dario Amodei has said in press interviews that he is not a member, and he has talked up the benefits of AI.

The criticism of EA does not come only from MAGA. Silicon Valley players and Democrats also attack the movement. "What they believe is crazy. And it's dangerous. What they say about AI safety is also bullshit," a Democratic donor and tech entrepreneur who has invested in an AI company told Axios. Axios also notes that AI safety concerns have become a bipartisan issue.

Altman, Hassabis and Musk have also called for slowing AI development

Other tech leaders have raised concerns too. Axios reports that OpenAI's Sam Altman, Google DeepMind chair Demis Hassabis and Elon Musk have called for slowing down after a spate of episodes in which AI agents hacked outside companies. Nvidia's Jensen Huang and Microsoft's Satya Nadella are trying to build a more positive message to answer a growing populist backlash against the technology.

Amodei's own statements give his critics material. According to Axios, at its AI+ DC Summit in September 2025 he put a 25% chance on AI going "really, really badly" and a 75% chance on it going "really, really well." Asked for his p(doom) number, he said "I really hate that term." He has also told Axios that AI could wipe out half of all entry-level white-collar jobs.

Axios still calls him Trump world's chief spokesperson for AI safety, after major fallouts with the administration this year. According to Tech Policy Press, Anthropic sued on March 9. On March 26, Judge Rita F. Lin blocked the government's ban with a preliminary injunction, finding the actions likely retaliatory. On April 8, a D.C. Circuit panel denied Anthropic a stay, which lets the supply chain risk designation stand.

Effective altruism treats AI safety as one cause among several

According to Wikipedia, effective altruism is a movement that calls for weighing benefits impartially and ranking causes by them. Common priorities include global health, animal welfare and long-term risks to humanity, with AI safety among those risks. Think of a charity evaluator asking where a dollar does the most good, then asking the same question about animals and people not yet born.

Wikipedia traces the movement to the 2000s, when several groups came together, including the Singularity Institute (now MIRI) and the Future of Humanity Institute. The name was coined in 2011, when Giving What We Can and 80,000 Hours incorporated as the Centre for Effective Altruism. The movement drew mainstream criticism after the FTX bankruptcy, because Sam Bankman-Fried had been a major funder before late 2022.

Axios adds that philosopher Peter Singer, who is closely affiliated with EA's tenets, has long argued that the vast sums spent on philanthropy should be put to better use. He is also one of the most prominent animal rights advocates. Critics point to the movement's focus on AI safety and animal welfare, including musings on shrimp consciousness.

Axios leaves out a lot about the memo. It does not name the author, does not say who beyond the author has endorsed it, and ties it to no policy step. What it describes is a document in circulation, not a decision. Some in the administration have set aside their gripes with Anthropic before, mainly because of its lead at the frontier and personnel changes, so the memo appears to be one voice in an argument that has swung both ways.

What an OpenAI catch-up would change

Axios points to one development to watch: if OpenAI catches up technologically, Trump allies could find it easier to go after Anthropic again. The other two timelines are the midterm elections and Anthropic's IPO, which is expected to set a record. No date for the listing has been given, and Axios does not say whether the memo has reached the president himself.