55% of US residents acknowledge the risk of extinction from AI

According to the results of a new public opinion study, 55% of the adult population in the US now considers artificial intelligence a potential existential threat. Data shows a sharp increase in public awareness of the risks associated with technological development, for comparison, at the end of 2022 this figure was only 7%. A 48 percentage point increase occurred over two years of active implementation of generative models in consumer products. The authors of the study note that the change in public perception directly affects the agenda in the field of AI state regulation. A high level of concern becomes a key factor that lawmakers consider when developing new safety standards and restrictions for developers of large neural network systems. The study highlights that discussions about "existential risk" (x-risk) have moved from narrow circles of experts and security researchers to the masses. This creates pressure on companies developing frontier models, requiring them to be more transparent and implement control mechanisms. The obtained figures reflect a shift in public discourse, where technological progress is increasingly compared to long-term risks for humanity. Such sociological measurements become an important indicator for assessing the readiness of society for further integration of autonomous systems into critical infrastructure.