Kokotajlo studied philosophy. He graduated from the University of Notre Dame with a BA in 2014 and went on to graduate school at the University of North Carolina at Chapel Hill, where he earned an MA in 2017. During his doctoral studies there he held the 2019–2020 Maynard Adams Fellowship for the Public Humanities. In 2021 he wrote a blog post called “What 2026 Looks Like”; in 2025 Kevin Roose wrote that a number of its predictions proved prescient.
In 2022 he joined OpenAI as a researcher in its governance division. He helped organize a group of employees who said the company had a secretive and reckless culture that was taking grave risks in the race to artificial general intelligence. When he resigned in 2024 he refused to sign OpenAI's non-disparagement clause, which could have cost him about $2 million in equity; in May 2024 he confirmed he kept his vested equity. In June 2024 he and other current and former AI company employees signed a letter calling for a “right to warn” about AI risks without reprisal.
He co-founded and leads the AI Futures Project, a nonprofit in Berkeley, California. In April 2025 it published AI 2027, a scenario in which AI automates software development and AI research and fully autonomous agents beat humans at “everything” around the end of 2027. In December 2025 the team pushed its estimate for full coding automation to the early 2030s, and in July 2026 it published AI 2040: Plan A, which recommends delaying superintelligence until 2040.
In September 2026 The Inference reported that Kokotajlo, who quit OpenAI in April 2024 after losing trust in its leadership, used a September 2026 appearance on Joe Rogan's show to challenge a system in which independent researchers need a company's goodwill to investigate its AI agents. The appearance came two months after the AI Futures Project published AI 2040: Plan A in July 2026.
Sources · written 16 October 2026