Yoshua Bengio
AI Pioneer & Safety Researcher
About
Yoshua Bengio is a professor at Université de Montréal and one of the pioneers of deep learning, sharing the 2018 Turing Award with Geoffrey Hinton and Yann LeCun. He founded Mila, the Quebec AI Institute, where he served as scientific director until 2025, and his work on neural language models, attention mechanisms, and generative adversarial networks helped lay the groundwork for modern AI. He chaired the first International AI Safety Report (2025) and co-founded LawZero to develop 'Scientist AI'—a non-agentic system designed to act as a guardrail against deceptive autonomous agents. Among the most-cited living scientists, he has become one of the most prominent voices urging caution as AI capabilities accelerate.
Key Contributions
- Pioneered neural probabilistic language models and distributed representations, early building blocks for modern language modeling
- Advanced representation learning through work on word embeddings, denoising autoencoders, and learning-to-learn methods
- Co-authored foundational work on attention for neural machine translation and on generative adversarial networks
- Founded Mila, turning Montréal into one of the world's major deep-learning research centers
- Shared the 2018 Turing Award with Hinton and LeCun for making deep learning a central AI paradigm
- Shifted from deep-learning booster to AI-safety advocate, chairing the International AI Safety Report and founding LawZero amid debate over how urgent frontier risks are
Videos & Interviews
The Catastrophic Risks of AI — and a Safer Path | Yoshua Bengio | TED
A TED talk laying out the most serious near-term and existential risks from increasingly capable AI systems, and outlining the technical and governance directions Bengio sees as a credible safer path.
View Details
Godfather of AI: How To Make Safe Superintelligent AI – Yoshua Bengio
Bengio explains his proposal for "Scientist AI"—a non-agentic, truthful system designed to serve as a guardrail against deceptive autonomous agents—and why he has grown more optimistic that safe superintelligence is technically achievable.
View Details
