Anthropic released Claude, its first public conversational AI assistant, initially through an API available to select partners and developers. The model was trained with the company's Constitutional AI approach, which aimed to steer behavior with a written set of principles rather than pure human feedback.

Claude arrived about three and a half months after ChatGPT and positioned itself as a safety-first alternative from a team founded by former OpenAI researchers. Early access went through a waitlist while Anthropic scaled up capacity.

The launch began the numbered Claude lineage that Anthropic has iterated ever since, from Claude 2 through the 3.x, 4.x, and 5.x generations.