Florian Tramèr

Florian Tramèr is an assistant professor of computer science at ETH Zurich. His research interests lie in Computer Security, Cryptography and Machine Learning security. In his current work, he studies the worst-case behavior of Deep Learning systems from an adversarial perspective, to understand and mitigate long-term threats to the safety and privacy of users.

AI2050 Project

AI systems like ChatGPT can be exploited as tireless, scalable assistants for cyberattacks, surveillance, and privacy violations. Unlike human hackers, AI doesn’t tire, can run thousands of operations simultaneously, and is cheap.

Tramèr’s project is building the first comprehensive public benchmarks to measure exactly how dangerous these capabilities are today and how they grow as AI improves. Their work spans three areas: measuring how AI can link anonymous online identities at scale, evaluating whether AI agents can conduct surveillance autonomously, and testing whether AI can break cryptographic systems. These measurements will inform AI safety policies, defensive tools, and responsible deployment decisions.