Professor Nicolas Papernot (ECE) has been named the 2025 recipient of the Steacie Prize, one of Canada’s most prestigious early career research awards, for his groundbreaking work advancing the security, privacy and trustworthiness of machine learning algorithms. The Steacie Prize is given annually to one engineer or scientist, 40 years of age or younger, who has made notable contributions to research in Canada.
Papernot is a pioneer in the development of adversarial examples in machine learning (ML). These are specially crafted inputs that force an ML model to make an incorrect prediction, also known as an “attack.” His research made clear that attacks against ML models were not just a hypothetical threat.Papernot also introduced several techniques that are now widely used to evaluate defences against attacks. This work has been so influential that the resulting research area now has a dedicated conference — the IEEE Conference on Secure and Trustworthy Machine Learning — which Papernot helped create and establish.
Papernot was recruited to join U of T Engineering and the Vector Institute in 2019 as part of Canada’s Pan-Canadian AI Strategy. Soon after, his group released a paper on “machine unlearning,” which was a milestone in defining how ML models could be made to forget some of their training data. Papernotbuilt on this research to develop a set of techniques for accounting and auditing in ML, allowing companies and institutions to provide verifiable privacy guarantees to their end users. For example, his group proposed a way for institutions to prove to a user that their ML models have “unlearned” the user’s data.
Papernot is a key contributor to the global discussion around the ethics and societal impact of ML. In 2020, he was invited to meet with the Privacy Commissioner of Canada to discuss the implications of ML on the privacy of individuals whose data is analyzed by governments and private companies. In 2024, he appeared before the House of Commons to comment on Bill C-27, which sought to regulate AI and data privacy. That same year, he chaired the Royal Society of Canada’s Task Force on Data Security and was named co-director of the Canadian AI Safety Institute Research Program at CIFAR.
In 2022, Papernot was named a Sloan Research Fellow and in 2023, he was elected to the Royal Society of Canada’s College of New Scholars, Artists and Scientists. His many other recognitions include the AI2050 Early Career Fellowship from Schmidt Sciences, and the Outstanding Early Career Researcher Award from the Association for Computing Machinery Special Interest Group on Security, Audit and Control.
“While still early in his career, Professor Nicolas Papernot has already made an extraordinary impact through his leadership in addressing the security and privacy challenges around machine learning and AI,” says U of T Engineering Dean Christopher Yip.
“On behalf of the faculty, I congratulate him on being recognized as one of Canada’s most promising science and engineering scholars.”
