Is Anthropic a cult?

Where the label comes from

Former OpenAI staff, among them Dario and Daniela Amodei, founded Anthropic in 2021 to focus on AI safety. Many early staff had links to the effective altruism movement, which some critics distrust.

Anthropic’s leaders write long essays about how AI could change the world, for better and for worse. To some readers, that sounds more like belief than business.

What Anthropic actually is

Anthropic is a public benefit corporation. It has investors and paying customers and competes with OpenAI and Google. Its staff are free to leave, and many speak in public.

It publishes its safety policies and research, which outside groups such as METR review and criticize.

Is Anthropic more ethical than OpenAI?

That depends on what you value, so compare what each company publishes and does. Both publish safety policies and reports on their models. Anthropic is a public benefit corporation with a Long-Term Benefit Trust that, according to its S-1, elects a majority of its board. OpenAI is controlled by a non-profit foundation.

Both companies have been criticized, by former employees, safety researchers and authors whose work was used for training.

Did Anthropic destroy books?

Yes, for scanning. Court records show that Anthropic bought millions of print books, cut off their bindings, scanned the pages and threw the paper away. In June 2025, a US judge ruled that this, and training on the books, was fair use.

The same judge ruled that downloading pirated books was not. Anthropic settled the authors’ class action for $1.5 billion, about $3,000 per book, without admitting liability.

Judge by actions

Read what Anthropic announces, then compare it with what independent testers and critics say. We stick to the facts and leave the judgement to you.