Who's who in AI
Public-record only. The work, not the gossip.
How to read this page
- Each entry is a two-sentence summary: affiliation, then the contribution they are publicly known for.
- Affiliation reflects best-effort public record as of June 2026. Check the linked profile for current status.
- Disagreements between people on this list (e.g. LeCun vs. Hinton on existential risk) are real and on the public record. We name them; we do not pick a side.
- If a name you expected is missing, that is not a judgment — the list is a starting set, not a ranking.
- Every link goes to a primary source: a personal page, a lab page, a Wikipedia entry, or a paper. We do not link to social media as a source.
The founding generation of deep learning
Three people share the 2018 Turing Award for the work that made the current wave of AI possible. Read their original papers, not just the commentary about them.
Geoffrey Hinton
en.wikipedia.org/wiki/Geoffrey_Hinton
Emeritus professor, University of Toronto. Co-author of the 1986 backpropagation paper with Rumelhart and Williams and the 2012 AlexNet paper with Krizhevsky and Sutskever; left Google in May 2023 to speak more freely about AI risk, and shared the 2024 Nobel Prize in Physics with John Hopfield for foundational work on artificial neural networks.
Yann LeCun
yann.lecun.com
Chief AI scientist at Meta and professor at NYU. Author of the convolutional neural network architecture (LeNet, 1989-1998) that underlies modern computer vision, and the most prominent public skeptic of near-term existential risk arguments, arguing instead for world-model approaches such as JEPA.
Yoshua Bengio
yoshuabengio.org
Professor at Université de Montréal and scientific director of Mila. Co-developed the foundational work on neural language models and attention mechanisms in the 2000s and 2010s, and chairs the International Scientific Report on the Safety of Advanced AI commissioned after the 2023 Bletchley summit.
Frontier-lab principals
The people running or co-founding the largest model labs as of June 2026. Where someone has publicly departed a lab, we mark that explicitly.
Researchers whose papers you will keep seeing cited
These are people whose specific technical contributions appear in nearly every modern ML curriculum. The link in each card is to a personal page or lab page that is more durable than any one paper.
Fei-Fei Li
profiles.stanford.edu/fei-fei-li
Professor at Stanford and co-director of the Stanford Institute for Human-Centered AI (HAI). Created ImageNet (2009), the dataset and benchmark that catalyzed the deep-learning revolution in computer vision, and co-founded World Labs (2024) on spatial intelligence.
Andrew Ng
andrewng.org
Founder of DeepLearning.AI and Landing AI; adjunct professor at Stanford. Co-founded Google Brain (2011), led Baidu AI Group (2014-2017), and built the original Stanford / Coursera machine-learning course that introduced millions of practitioners to the field.
Stuart Russell
people.eecs.berkeley.edu/~russell
Professor at UC Berkeley and co-author with Peter Norvig of 'Artificial Intelligence: A Modern Approach,' the standard AI textbook. Founded the Center for Human-Compatible AI (CHAI) at Berkeley; published 'Human Compatible' (2019) on the alignment problem and has briefed the UN on autonomous weapons.
Percy Liang
cs.stanford.edu/~pliang
Associate professor at Stanford and director of the Center for Research on Foundation Models (CRFM). Lead author of the 2021 'On the Opportunities and Risks of Foundation Models' report and of the HELM benchmark suite for holistic evaluation of language models.
Chelsea Finn
ai.stanford.edu/~cbfinn
Assistant professor at Stanford and former research scientist at Google Brain. Best known for the Model-Agnostic Meta-Learning (MAML) algorithm and for foundational work on robot learning from demonstration and language-conditioned manipulation.
Anca Dragan
people.eecs.berkeley.edu/~anca
Associate professor at UC Berkeley and (as of 2024) head of AI safety and alignment at Google DeepMind. Research focuses on human-robot interaction, inverse reward design, and how robots should communicate intent legibly to humans.
Sara Hooker
sarahooker.me
Vice president of research at Cohere and head of Cohere For AI. Author of 'The Hardware Lottery' (2020), a widely-cited essay on how hardware constraints shape which ideas succeed in ML, and a prominent voice on open-science and multilingual model evaluation.
Paul Christiano
paulfchristiano.com
Head of the U.S. AI Safety Institute (within NIST) as of 2024. Previously ran the Alignment Research Center (ARC) and led the language model alignment team at OpenAI; foundational author of the 2017 'Deep Reinforcement Learning from Human Preferences' paper that underlies RLHF.
Public commentators on AI safety and risk
Founders building on top of frontier models
These are founders whose companies depend on, or compete with, the frontier labs. Several have a history at those labs.
Mustafa Suleyman
en.wikipedia.org/wiki/Mustafa_Suleyman
CEO of Microsoft AI (since March 2024). Co-founded DeepMind (2010) with Demis Hassabis and Shane Legg; co-founded Inflection AI in 2022; author of 'The Coming Wave' (2023) on the governance challenges of AI and synthetic biology.
Aravind Srinivas
perplexity.ai
Co-founder and CEO of Perplexity AI. Previously a research scientist at OpenAI and Google; the company shipped one of the earliest production answer-engine products built on top of frontier LLMs combined with retrieval.
Emad Mostaque
en.wikipedia.org/wiki/Emad_Mostaque
Founder and former CEO of Stability AI, the company that released Stable Diffusion in August 2022; resigned as CEO in March 2024. His subsequent ventures are publicly announced but technical output is limited as of June 2026 — check primary sources.
Bryan Catanzaro
ctnzr.io
Vice president of applied deep-learning research at NVIDIA. Led the team that built Megatron-LM and several of the techniques (tensor parallelism, large-batch training tricks) that make trillion-parameter training tractable on NVIDIA hardware.
The hardware behind the software
Models do not train without chips. These two companies are the dominant suppliers of AI accelerators as of June 2026.
Selected primary papers worth reading directly
If you are going to read three papers from the people above, read these. arXiv IDs verified against the public arXiv index.
Public letters and statements worth knowing about
Three documents that signaled different positions in the public debate. Read each before forming an opinion about who agrees with whom.
March 2023
Future of Life Institute 'Pause Giant AI Experiments' open letter
Called for a six-month pause on training systems more powerful than GPT-4. Signed by Bengio, Russell, and thousands of others; not signed by most frontier-lab leadership. The original letter remains hosted on futureoflife.org.
May 2023
Center for AI Safety one-sentence statement on extinction risk
Single sentence equating AI extinction risk with pandemic and nuclear risk. Signed by Hinton, Bengio, Hassabis, Altman, Amodei, Russell, and hundreds more. LeCun and Andrew Ng publicly declined to sign and have argued against this framing in the months since.
November 2023
Bletchley Declaration
First international government-level statement on frontier AI safety, signed at the UK's AI Safety Summit by the U.S., U.K., E.U., China, and 25 other governments. Led to the creation of the U.K. AI Safety Institute and the U.S. AI Safety Institute, the latter now headed by Paul Christiano.
A short honesty note about what this page does not do
This page is a starting index. It does not rank people, it does not score labs against each other, and it does not include private information about anyone listed. Several people on this list have publicly disagreed with each other in strong terms about whether current AI systems pose existential risk, about whether scaling will continue to yield capabilities, and about whether open-source release is a net good. We have tried to present the public positions accurately without endorsing any of them. If we have misstated a current affiliation — for instance, if someone has moved labs since June 2026 — please treat the linked source as authoritative over our summary. The point of this page is to send you to people's own words, not to replace them.
Sources
- [01]
Geoffrey Hinton received the 2018 Turing Award and the 2024 Nobel Prize in Physics (with John Hopfield) for foundational work on artificial neural networks.
en.wikipedia.org/wiki/Geoffrey_Hinton
- [02]
Yann LeCun's personal page documents his role as chief AI scientist at Meta and his work on convolutional networks (LeNet) starting in 1989.
yann.lecun.com
- [03]
Yoshua Bengio is professor at Université de Montréal and scientific director of Mila, and chairs the International Scientific Report on the Safety of Advanced AI.
yoshuabengio.org
- [04]
Demis Hassabis co-founded DeepMind in 2010 and leads Google DeepMind.
deepmind.google/about/
- [05]
Hassabis and John Jumper shared the 2024 Nobel Prize in Chemistry with David Baker for protein structure prediction.
nobelprize.org/prizes/chemistry/2024/
- [06]
The 2017 'Attention Is All You Need' paper introduced the transformer architecture.
arxiv.org/abs/1706.03762
- [07]
The 2020 'Language Models are Few-Shot Learners' paper introduced GPT-3.
arxiv.org/abs/2005.14165
- [08]
Christiano, Leike, Brown et al. (2017) introduced learning from human preferences, the foundational RLHF technique.
arxiv.org/abs/1706.03741
- [09]
Anthropic's 2022 Constitutional AI paper describes harmlessness training from AI feedback.
arxiv.org/abs/2212.08073
- [10]
Stanford CRFM's 2021 report 'On the Opportunities and Risks of Foundation Models' (Bommasani et al.) coined the foundation-models framing.
arxiv.org/abs/2108.07258
- [11]
Sara Hooker's 2020 essay 'The Hardware Lottery' argues that hardware availability shapes which ML ideas succeed.
arxiv.org/abs/2009.06489
- [12]
He, Zhang, Ren, Sun (2015) introduced ResNet / deep residual learning for image recognition.
arxiv.org/abs/1512.03385
- [13]
Jumper, Evans, Pritzel et al. published AlphaFold 2 in Nature 596 (2021).
nature.com/articles/s41586-021-03819-2
- [14]
The Center for AI Safety May 2023 one-sentence statement on extinction risk was signed by Hinton, Bengio, Hassabis, Altman, Amodei, Russell, and hundreds of others.
safe.ai/statement-on-ai-risk
- [15]
The March 2023 Future of Life Institute open letter called for a six-month pause on training systems more powerful than GPT-4.
futureoflife.org/open-letter/pause-giant-ai-experiments/
- [16]
The November 2023 Bletchley Declaration was signed by the U.S., U.K., E.U., China, and 25 other governments at the U.K. AI Safety Summit.
gov.uk/government/publications/ai-safety-summit-2023-the-bletchley-declaration
- [17]
The U.S. AI Safety Institute is housed within NIST; Paul Christiano was named head of AI safety in 2024.
nist.gov/aisi
- [18]
Anthropic was co-founded in 2021 by Dario Amodei, Daniela Amodei, and several former OpenAI colleagues.
anthropic.com/company
- [19]
Sam Altman was briefly removed and reinstated as CEO of OpenAI in November 2023.
openai.com/blog/sam-altman-returns
- [20]
Ilya Sutskever co-founded Safe Superintelligence Inc. (SSI) in June 2024 after leaving OpenAI.
ssi.inc
- [21]
Fei-Fei Li is co-director of the Stanford Institute for Human-Centered AI and created the ImageNet dataset in 2009.
hai.stanford.edu/people/fei-fei-li
- [22]
Percy Liang directs the Stanford Center for Research on Foundation Models (CRFM).
crfm.stanford.edu
- [23]
Stuart Russell founded the Center for Human-Compatible AI at UC Berkeley and published 'Human Compatible' in 2019.
humancompatible.ai
- [24]
Helen Toner is director of strategy at Georgetown's Center for Security and Emerging Technology (CSET).
cset.georgetown.edu/staff/helen-toner/
- [25]
Chris Olah leads the long-running Transformer Circuits / mechanistic interpretability work at Anthropic.
transformer-circuits.pub