A test by CORRECTIV.Schweiz shows how easily AI chatbots can generate convincing fake news in the style of established media outlets. Professor Florian Tramèr, head of the Security and Privacy Research Group at ETH Zurich, highlights the risks of such misuse, particularly when fake content uses trusted media brands to appear credible. Read article in German.
D-INFK alumnus Severin Hacker co-founded the language-learning company Duolingo in 2011 and has served as CTO since its inception. He announced on LinkedIn that he has been "shifting into a more hands on role" at the company and his investor relations page indicates that he is no longer CTO.
Professor Martin Vechev, head of the Secure, Reliable, and Intelligent Systems Lab (SRI Lab), and his team investigated the impact of LLM watermarks on medical AI systems. Their study shows that watermarking techniques, while enabling traceability of AI-generated content, can negatively affect clinical reasoning quality and introduce new failure modes.