Survey evidence on public support for AI safety oversight

Abstract

A number of AI safety concerns are being increasingly discussed by experts, including misinformation, invasion of privacy, job displacement, and criminal misuse. Two exploratory studies conducted in Germany and Spain (combined n = 2864) provide evidence that the general public largely supports strict oversight over safety of commercial artificial intelligence research. Among the factors that are associated with preferences for strict oversight are age, anticipated job displacement, innovativeness, and risk, time and altruistic preferences.