Anthropic

Anthropic is an American AI company founded in 2021 by a group of former OpenAI researchers, including siblings Dario and Daniela Amodei. It builds the Claude family of AI models and positions safety and reliability as central to its mission.

The company has grown quickly into one of the handful of “frontier” labs whose models sit at the leading edge of capability. To train and run them, Anthropic has signed enormous compute deals, including reserving city-scale amounts of electrical power and custom chips years in advance.

Anthropic also invests heavily in research on how to make powerful models behave predictably and resist misuse, and it has attracted senior scientists from rival labs. For readers of this site, Anthropic is relevant both as a model provider and as a frequent subject of debates about AI safety, cost and governance.