Yoshua Bengio

Regular research member

ORCID Github Google Scholar Page personnelle

Thematic groups

Recognized worldwide as one of the leading experts in artificial intelligence, Yoshua Bengio is most known for his pioneering work in deep learning, earning him the 2018 A.M. Turing Award, “the Nobel Prize of Computing,” with Geoffrey Hinton and Yann LeCun. He is Full Professor at Université de Montréal, and the Founder and Scientific Director of Mila – Quebec AI Institute. He co-directs the CIFAR Learning in Machines & Brains program as Senior Fellow and acts as Scientific Director of IVADO and holds a Canada CIFAR AI Chair. In 2023, he was awarded the prestigious Gerhard Herzberg Canada Gold Medal for Science and Engineering, the country's most significant scientific prize. He is the computer scientist with the highest h-index citation metric and ranked third among all disciplines by scientific impact according to the Stanford bibliometric study. He is a Fellow of both the Royal Society of London and Canada, ACM Fellow, Knight of the Legion of Honor of France, Officer of the Order of Canada, Member of the UN’s Scientific Advisory Board for Independent Advice on Breakthroughs in Science and Technology and Scientific advisor for the UK AI Safety Institute (AISI). Concerned about the social impact of AI, he actively contributed to the Montreal Declaration for the Responsible Development of Artificial Intelligence and devotes himself to reducing the catastrophic risks of future AI and devotes himself to reducing the catastrophic risks of future AI.

Publications

Open Problems in Technical AI Governance

Yoshua Bengio

AI progress is creating a growing range of risks and opportunities, but it is often unclear how they should be navigated. In many cases, the barriers and uncertainties faced are at least partly technical. Technical AI governance, referring to technical analysis and tools for supporting the effective governance of AI, s

Government interventions to avert future catastrophic AI risks

Yoshua Bengio

Transcription révisée du témoignage de Yoshua Bengio en juillet 2023 devant la sous-commission du Sénat américain sur la vie privée, la technologie et le droit, qui s'est réunie sur le thème de la surveillance de l'IA. Il plaide en faveur de la prudence et des interventions gouvernementales en matière de réglementati ...

Government Interventions to Avert Future Catastrophic AI Risks

Yoshua Bengio

This essay is a revised transcription of Yoshua Bengio’s July 2023 testimony in front of the U.S. Senate Subcommittee on Privacy, Technology, and the Law meeting on the topic of oversight of AI. It argues for caution and government interventions in regulation and research investments to mitigate the potentially catastr

Managing extreme AI risks amid rapid progress

Yoshua Bengio

Artificial intelligence (AI) is progressing rapidly, and companies are shifting their focus to developing generalist AI systems that can autonomously act and pursue goals. Increases in capabilities and autonomy may soon massively amplify AI’s impact, with risks that include large-scale social harms, malicious uses, and

Regulating advanced artificial agents

Yoshua Bengio, Michael K.  Cohen, Noam  Kolt, Gillian K.  Hadfield, Stuart Russell 

Technical experts and policy-makers have increasingly emphasized the need to address extinction risk from artificial intelligence (AI) systems that might circumvent safeguards and thwart attempts to control them (1). Reinforcement learning (RL) agents that plan over a long time horizon far more effectively than humans

Interdisciplinary Dialogues: The Major Risks of Generative AI

Yoshua Bengio, Caroline Lequesne, Hugo Loiseau, Jocelyn Maclure, Juliette Powell, Sonja Solomun, Lyse Langlois

In an exciting series of Interdisciplinary Dialogues on the societal impacts of AI, we invite a guest speaker and panellists from the fields of science and engineering, health and humanities and social sciences to discuss the advances, challenges and opportunities raised by AI.The first dialogue in this series began wi ...

Computing Power and the Governance of Artificial Intelligence

Yoshua Bengio, Gillian K.  Hadfield

Computing power, or "compute," is crucial for the development and deployment of artificial intelligence (AI) capabilities. As a result, governments and companies have started to leverage compute as a means to govern AI. For example, governments are investing in domestic compute capacity, controlling the flow of compu ...

Generative AI models should include detection mechanisms as a condition for public release

Alistair Knott, Dino Pedreschi, Raja Chatila, Tapabrata Chakraborti, Susan Leavy, Ricardo Baeza‑Yates, David Eyers, Andrew Trotman, Paul D. Teal, Przemyslaw Biecek, Stuart Russell , Yoshua Bengio

The new wave of ‘foundation models’—general-purpose generative AI models, for production of text (e.g., ChatGPT) or images (e.g., MidJourney)—represent a dramatic advance in the state of the art for AI. But their use also introduces a range of new risks, which has prompted an ongoing conversation about possible regulat

AI and Catastrophic Risk

Yoshua Bengio

Since OpenAI's release of the very large language models Chat-GPT and GPT-4, the potential dangers of AI have garnered widespread public attention. In this essay, the author reviews the threats to democracy posed by the possibility of "rogue AIs," dangerous and powerful AIs that would execute harmful goals, irrespectiv

International Institutions for Advanced AI

Yoshua Bengio

International institutions may have an important role to play in ensuring advanced AI systems benefit humanity. International collaborations can unlock AI's ability to further sustainable development, and coordination of regulatory efforts can reduce obstacles to innovation and the spread of benefits. Conversely, the p

FAQ on catastrophic AI risks

Yoshua Bengio

I have been hearing many arguments from different people regarding catastrophic AI risks. I wanted to clarify these arguments, first for myself, because I would really like to be convinced that we need not worry. However, sharing them and opening up the discussion more broadly might also be useful.

Model evaluation for extreme risks

Yoshua Bengio

Current approaches to building general-purpose AI systems tend to produce systems with both beneficial and harmful capabilities. Further progress in AI development could lead to capabilities that pose extreme risks, such as offensive cyber capabilities or strong manipulation skills. We explain why model evaluation is c

How Rogue AIs may Arise

Yoshua Bengio

he rise of powerful AI dialogue systems in recent months has precipitated debates about AI risks of all kinds, which hopefully will yield an acceleration of governance and regulatory frameworks. Although there is a general consensus around the need to regulate AI to protect the public from harm due to discrimination an

Innovation ecosystems for socially beneficial AI

Yoshua Bengio, Allison Cohen, Benjamin Prud’homme, Amanda Leal, Noah Oder

Governments have been preoccupied with the disruptive potential that AI technology confers, wanting to either maximize the creation of economic growth or minimize the risk of rights-related violations. Consequently, governments and international institutions have focused their efforts either on funding the AI industry

Events

Conference-discussion: Anticipating AI's major Risks for Society

Panel | Hybrid

Mila, OBVIA and IVADO invite their academic communities to join an exclusive discussion with Yoshua Bengio and an interdisciplinary panel of experts aimed at engaging a reflection on the major risks of AI. For several months, researchers in many fields have feared the cata ...

News

Auditions publiques à l'Assemblée nationale sur les outils de notification de traçage numérique

Plusieurs membres chercheurs de l’Observatoire international sur les impacts sociétaux de l’IA et du numérique (OBVIA) ont participé cette semaine aux consultations particulières et auditions publiques au sujet d'outils technologiques de notification des contacts. Ils ont ...

L'OBVIA s'associe à Avant Garde

Cette initiative du CIRANO, en collaboration avec l'OBVIA et IVADO, vise à rassembler et accompagner 45 jeunes leaders issus de différents horizons dans leur développement professionnel en leur proposant notamment des conférences et ateliers s’appuyant sur la recherche. Le ...

Media interventions

Les risques croissent plus vite que les solutions

L’intelligence artificielle continue de progresser en 2026, mais pas toujours dans la direction souhaitée. D’un côté, des agents d’IA de plus en plus autonomes se sont dotés de leur propre forum web. De l’autre, des IA insuffisamment testées avant d’être publiées accroissent les risques d’une application néfaste de la technologie. Et entre les deux, les solutions pour éviter les dérapages semblent de moins en moins nombreuses…

Media

La Presse

Entre occasion et désastre, préviennent des experts

Manque de main-d’œuvre expérimentée, hégémonie de la Chine et des États-Unis et prix de l’électricité en forte hausse font partie des effets économiques potentiellement désastreux de l’intelligence artificielle, ont prévenu lundi à Montréal deux sommités. Mais elles y voient aussi un potentiel important.

Media

La Presse

Les promesses et les dangers de l’IA, selon Acemoğlu et Bengio

L’intelligence artificielle vient avec des promesses, mais aussi avec de graves dangers, préviennent deux sommités mondiales, Daron Acemoğlu et Yoshua Bengio, qui s’inquiètent qu’on ait autant de mal à en prendre conscience.

Media

Le Devoir

Stay up to date with the latest news!

Receive updates from the community, upcoming events, and research opportunities.