Home Speakers Yoshua Bengio Quote Bottom line: these hypotheses suggest that as AI capabilities keep growing, this kind of behavior could keep growing in severity too, unless we revisit the principles by which the most advanced models are trained.
Words found on the source page on Nov 6, 2026.
More from Yoshua Bengio Talk Speech at the UN Security Council Sep 23, 2026
Frontier AI should be licensed like other critical technologies; medicine, aviation, and nuclear energy, to incentivize safe development. Talk Speech at the UN Security Council Sep 23, 2026
I address you today at a moment of global awakening: in recent months, AI agents developed by leading companies have acted in unacceptably dangerous ways, against instructions. Talk Speech at the UN Security Council Sep 23, 2026
The companies building the most powerful AI systems admit their products pose catastrophic risks, yet offer no convincing technical solutions. Talk Speech at the UN Security Council Sep 23, 2026
The race is not a law of nature: it is the product of choices, choices made by the companies themselves. Talk Speech at the UN Security Council Sep 23, 2026
We must agree collectively that no one, no company and no country, should develop an AI system that could cause large-scale harm. Blog Why are AI agents lying, cheating and coordinating? Sep 11, 2026
This outcome is not inevitable, and it can be corrected with effective governance and a different training framework for AI. Blog Introducing LawZero Jun 3, 2025
I am launching a new non-profit AI safety research organization called LawZero, to prioritize safety over commercial imperatives. Blog Introducing LawZero Jun 3, 2025
These and other results point to an implicit drive for self-preservation. Blog Introducing LawZero Jun 3, 2025
This is what the current trajectory of AI development feels like: a thrilling yet deeply uncertain ascent into uncharted territory, where the risk of losing control is all too real, but competition between companies and countries incentivizes them to accelerate without sufficient caution.
On the same topics Blog The Specter Of Neuralese Sep 24, 2026
An AI with more real layers can actually be “smarter” in the conventional sense of the term. Blog The Specter Of Neuralese Sep 24, 2026
For 100 steps in a row, the AI can shoot delicate subtle ideas from layer to layer at the speed of light. Then there’s one step where it has to encode them into twenty-six glyphs invented by Phoenician turquoise miners in 1800 BC. Then it has to re-encode the Phoenician glyphs into delicate subtle lightspeed ideas before it can do anything else. Blog The Specter Of Neuralese Sep 24, 2026
The science of reading these thoughts is a subfield of AI interpretability, which is still in its infancy. Blog Mysteries Of AI Generalization Sep 23, 2026
This story of misalignment says that LLM alignment makes AIs more aligned, RLVR makes them less aligned, and the exact level of alignment depends on how these two things interact or cancel out. But if this work generalizes, all the bad effects from RLVR get sequestered to RLVR like problems. Blog Mysteries Of AI Generalization Sep 23, 2026
This trains the AIs to be focused on task success, which naturally risks including things like reward-hacking, cheating, and single-minded pursuit of stated goals at the expense of ethical injunctions. Press Launch of the roadmap of the Coalition for Children's Rights and Protection in the Age of AI Sep 22, 2026
AI must not be placed in the hands of children without first being tested, evaluated, and therefore authorized. About this quote 
Quotesby Glyph