Nils Lukas
Assistant Professor of Machine Learning
Research interests
Professor Lukas has a background in machine learning (ML), cryptography, security, and privacy, and is interested in researching trustworthy ML systems in the presence of untrustworthy entities. This field of research includes a range of topics such as controlling the potential misuse of ML systems, enhancing the reliability of generated outputs, studying aligned and explainable ML systems and using confidential computing methods. Most recently, he has been active in developing provably robust watermarking methods to detect ML-generated content, auditing differentially private language models, and inferring confidential user data with ML systems without revealing it to the ML service provider. Email
- Ph.D. in trustworthy machine learning from the University of Waterloo in Canada.
- Master of Science on privacy-enhancing technologies from the RWTH-Aachen in Germany.
- Bachelor of Science in computer science from the RWTH-Aachen in Germany.
- Mathematics Doctoral Prize’s Top Honor, 2024.
- Governor’s General Gold Medal Finalist in Canada, 2024.
- Distinguished contribution award at Microsoft MLADS, 2023.
- Spotlight Award, ICLR 2021.
- Nils Lukas, Abdulrahman Diaa, Lucas Fenaux, Florian Kerschbaum: “Leveraging Optimization for Adaptive Attacks on Image Watermarks”, ICLR 2024, The Twelfth International Conference on Learning Representations.
- Benjamin Schneider, Nils Lukas, Florian Kerschbaum: “Universal Backdoor Attacks”, ICLR 2024, The Twelfth International Conference on Learning Representations.
- Nils Lukas, Ahmed Salem, Robert Sim, Shruti Tople, Lukas Wutschitz, Santiago Zanella-Béguelin: “Analyzing Leakage of Personally Identifiable Information in Language Models”, IEEE S&P 2023, The 2023 Symposium on Security and Privacy.
- Nils Lukas and Florian Kerschbaum: “PTW: Pivotal Tuning Watermarking for Pre-Trained Image Generators”, USENIX 2023, The 33rd USENIX Security Symposium.
- Nils Lukas, Edward Jiang, Xinda Li, Florian Kerschbaum: “SoK: How Robust is Image Classification Deep Neural Network Watermarking?”, IEEE S&P 2022, The 2022 Symposium on Security and Privacy.
- Nils Lukas, Yuxuan Zhang, Florian Kerschbaum: “Deep Neural Network Fingerprinting by Conferrable Adversarial Examples”, ICLR 2021, The Ninth International Conference on Learning Representations.