Anthropic CEO on OpenAI Exit: Focus on AI Safety

Anthropic CEO Dario Amodei has discussed his departure from OpenAI. He explains that he left to build a competitor focused on AI safety. Amodei joined OpenAI in 2016 and became VP of Research. He left in late 2020, along with about a dozen other researchers, including his sister. The core tension, according to Amodei, was a contradiction between OpenAI's stated commitment to safety and the direction its leadership was taking the company. He then founded Anthropic in early 2021. The company aims to build AI systems that are both capable and genuinely safe. Anthropic has since developed the Claude family of models, which compete with OpenAI's GPT products. Amodei cites two main reasons for the split: a belief in scaling laws for AI capability and the conviction that safety research must be integrated from the start. For investors, tokenized pre-IPO instruments tied to Anthropic have circulated on the Solana blockchain. However, Anthropic warned in May 2026 that unauthorized transfers of these tokens would be voided, leading to a 30% drop in their value. This means holders of these instruments have essentially no legal protection.