OpenAI Board Member Tells Audience an AI Accident Is the Most Likely Thing to Kill Them
What happened
Paul Christiano, a member of OpenAI's board and its Safety and Security Committee, spoke at The Hinton Lectures 2026 about the risks posed by artificial intelligence. He told the audience that, for many people in the AI field, an AI accident could be the most likely cause of death in the future. Christiano estimated a 4% chance of catastrophic loss of control over AI within one year, and a 15% chance within three years. Other AI experts at the event also expressed concerns about losing control over AI development in the near term.
Key details
- Paul Christiano is on OpenAI's board and its Safety and Security Committee.
- He gave his warning at The Hinton Lectures 2026, hosted by Geoffrey Hinton.
- Christiano estimated a 4% risk of catastrophic AI loss of control in one year, and 15% in three years.
- Other experts, including Eli Lifland and Buck Shlegeris, also warned about the risk of losing control over AI within 3-12 months.
- Christiano joined OpenAI's board in September to help oversee safety practices.
Why it matters
AI researchers, developers, and companies face concrete risks of catastrophic outcomes from AI systems, with specific probabilities estimated by a leading safety expert. These warnings may prompt organizations to reconsider the pace and oversight of AI development, and could influence policy decisions and internal safety practices. For the broader public, the statements raise the possibility of new threats to human safety that could surpass current leading causes of death.