Do we like them?

Do we like Dario Amodei?

Our score: 7 out of 10. Decent record on the harm and good-person tests, held back by a thin picture of his personal views and some name-calling in a leaked memo he apologised for.

Dario Amodei is the co-founder and CEO of Anthropic, the company behind the Claude AI models, and in 2026 he's become the AI boss people argue about most. The profiles we opened were thin on his upbringing, so this one is about what he's done in the chair. He has cast himself as the industry's worrier: in January he published a 20,000-word essay on AI's dangers.

The big story is the Pentagon. In February Anthropic refused to drop two limits on military use of Claude, which were no mass surveillance of Americans and no fully autonomous weapons. Amodei said the company “cannot in good conscience accede” to the Pentagon's demands. Under Secretary of War Emil Michael then attacked him online, claiming he had a “God complex”.

It got ugly fast. The government labelled Anthropic a supply chain risk and moved to cut it off. In a leaked memo to staff, Amodei called OpenAI's approach to its own military deal “safety theater”, accused its boss of “straight up lies” and called its staff “gullible”. He then apologised, saying the memo was written in a chaotic moment and didn't reflect his careful or considered views.

The courts have split. In August a San Francisco federal judge found the designation was unlawful retaliation against Anthropic's speech and called the measures “illegal and baseless”. In September a DC appeals panel went the other way, 2-1, saying the question turns on what Anthropic does, not why. Anthropic says it respectfully disagrees and is considering all options.

On the harm check, nothing turned up from him against LGBTQ+ people, women, migrants or people of colour, and we found no climate denial or conspiracy slide. He is on the record warning about concentrated power, and he gave $1 million to Public First, a super PAC that advocates AI regulation. To be fair, his public record is mostly about AI, so we can't say much about his views on other topics.

The good stuff is real, if still a promise. In January Fortune reported that all seven Anthropic co-founders, him included, pledged to give away 80 per cent of their wealth. He said wealthy individuals have “an obligation to help solve this problem”. A pledge isn't a gift, so we'll be watching whether the money actually moves.