Isabelle Guyon
Director and Research Scientist, Google DeepMind, Professor of Artificial Intelligence, University Paris-Saclay
Isabelle Guyon has been called the Joan of Arc of AI. She always seeks new "battles" of research. In the 1990s, while many turned to feedforward neural networks, she championed kernel methods and co-developed one of the most iconic algorithms in machine learning: Support Vector Machines. When kernel methods peaked in the early 2000s, she pivoted to causality – then an emerging area that has since drawn widespread interest.
“The weakest link is data.”
More recently, her focus has shifted to data-centric modeling. In her NeurIPS 2022 invited talk, she remarked, “the weakest link is data,” advocating not for more data, but for better data. Guyon is also a strong proponent of competitions to drive machine learning progress. She serves as President of ChaLearn, a non-profit fostering such challenges, and as Community Lead of Codalab, the most-used ML competition platform in 2025.
A leading figure in European AI, Guyon is Director and Research Scientist at Google DeepMind. She also holds a Chair in Artificial Intelligence and conducts research with INRIA at the University Paris-Saclay. Earlier in her career, she earned her PhD in Paris, spent seven years at Bell Labs, and founded the consulting firm Clopinet in Berkeley, USA. Guyon’s many accolades include the 2020 BBVA Award and the 2023 Ten-Year Technical Impact Award from ICMI. She has also been named to the French Academy of Technologies.