ONE IN FIVE AI RESEARCHERS FORESEE EXTINCTION RISK

The survey reveals significant anxiety within the AI research community about existential risks. Prominent researchers from major organizations have publicly flagged the threat: Anthropic's Jacob Coxon ignited debate on social media, OpenAI's Daniel Selsam described a "ticking time bomb," and a former DeepMind researcher warned AI could pose lethal risks. The 18 percent average masks variation in individual estimates, but the consensus points toward genuine concern among those building advanced AI systems. Notably, this figure represents baseline sentiment in 2024, with researchers indicating the probability estimate has been climbing. The growing acknowledgment of existential risk from AI among researchers contrasts with limited public discussion of the topic. As AI capabilities accelerate, the gap between researcher warnings and public awareness remains significant.