Anthropic was founded in 2021 by Dario Amodei, Daniela Amodei, Paul Christiano, and Tom Brown, who were former senior researchers at OpenAI. The founders left OpenAI due to disagreements about the company's direction and safety protocols, establishing Anthropic with a focus on AI safety and beneficial AI development.
The company quickly gained recognition for its research in AI alignment and safety, developing the concept of "Constitutional AI" - AI systems designed with explicit principles and constraints to ensure safe behavior. Anthropic's research papers on AI safety and alignment have been influential in the AI community.
Throughout 2022 and 2023, Anthropic developed the Claude family of AI models, which gained recognition for their helpfulness, safety features, and reduced likelihood of generating harmful content. The company's models competed with OpenAI's GPT series and other AI systems in the rapidly growing AI market.
In 2023, Amazon announced a significant investment in Anthropic, followed by additional funding rounds from Google, Salesforce Ventures, and other investors. This funding allowed Anthropic to scale up its research and development while maintaining its focus on AI safety and beneficial AI.
In 2025, Anthropic continued its innovation with the release of Claude Sonnet 4.5, which set new benchmark records in coding, reasoning, and computer use. The company also released the Claude Agent SDK for building capable AI agents.
February 2026 marked a major milestone for Anthropic with the introduction of Claude Opus 4.6, the company's smartest model. The same month, Anthropic raised $30 billion in Series G funding at a $380 billion valuation, with run-rate revenue reaching $14 billion. The company also partnered with NASA, with Claude assisting the Perseverance rover's drive on Mars.