18
AI Robustness Challenges and Solutions
- Topic: Emerging Technologies
- Time: 18.02.2025, 13:30 to 14:00
- Location: Tech Stage
- Language:
Large-scale AI language models (LLMs) and other generative AI are commoditising and automating intelligence. This could lead to computer malware changing. In this presentation, examples of possible "Malware 2.0" will be shown and their functions explained. It will also argue that advances in generative AI are more likely to benefit attackers than defenders. The discussion will therefore focus on how security strategies need to adapt in order to keep pace with developments in generative AI.
Speaker
Prof. Dr. Florian Tramèr
Professor Dr. Florian Tramèr is Assistant Professor of Computer Science at ETH Zurich and heads the SPY laboratory. As a member of the Institute for Information Security, the ZISC and an associate lecturer at the ETHZ AI Centre, he conducts research in the fields of computer security, machine learning and cryptography. He studies the worst-case behaviour of deep learning systems to understand and mitigate threats to user security and privacy. He received his PhD from Stanford University and spent a year at Google Brain.