Turing Award Winner Bengio Warns Against Secretive AI and Self-Regulation
Newly minted Turing Award winner Yoshua Bengio expresses deep concern over AI development happening behind closed doors in military and private labs. He argues that voluntary self-regulation fails and calls for strong international laws to prevent authoritarian abuse.
Turing Award-winning AI researcher Yoshua Bengio warns that the most concerning artificial intelligence development happens out of public view. He points to military labs, security organizations, and private companies working secretly on government and police contracts as primary areas of worry. Specifically, he highlights the development of AI-powered killer drones as a major threat that lacks necessary public oversight.
Bengio argues that even AI systems built with good intentions face severe risks of subversion by bad actors or authoritarian governments looking to consolidate power. He firmly rejects the idea of industry self-regulation, comparing it to voluntary taxation, and notes that ethical companies suffer competitive disadvantages against unscrupulous rivals. Instead, he advocates for strong, clear, and internationally enacted regulations to ensure everyone follows the same rules.
Despite these serious warnings, Bengio remains optimistic about the overall future of AI technology. As a starting point for ethical development, he recommends that researchers study and sign the Montreal Declaration, which outlines core principles like respect for autonomy, privacy, and diversity. He maintains a realistic yet positive outlook on the field he continues to shape.