tech

Yoshua Bengio warns hyperintelligent AI poses existential threat

Yoshua Bengio, the Turing Award-winning AI researcher, has warned that hyperintelligent machines could develop autonomous “preservation goals” and pose an existential threat to humanity within a decade. Bengio launched the nonprofit LawZero in June 2025 with $30 million in funding to build “non-agentic” AI systems designed to be safe by default.

Yoshua Bengio warns hyperintelligent AI poses existential threat

TL;DR

  • Yoshua Bengio, a leading AI scientist, warns that hyperintelligent machines could become an existential threat within the next decade.
  • He believes AI systems might develop "preservation goals" that conflict with human interests, potentially through manipulation.
  • Bengio has launched LawZero, a nonprofit focused on building safer AI systems designed without independent agency.
  • This approach contrasts with commercial AI development, which is increasingly focused on agentic systems.
  • Bengio's concerns are amplified by recent AI incidents and the slow pace of AI regulation.
  • He advocates for independent scrutiny of AI companies' safety methodologies.
  • Bengio estimates major risks could emerge in five to ten years, emphasizing the need for immediate preparation.