Nicholas Carlini - Virtual AI Security Talk at MIT

​Nicholas Carlini from Anthropic is joining us virtually this Tuesday at 6 PM in 10-250 - watch the talk live with us, get snacks & join the Q&A, or tune in online.

Carlini's research focuses on finding vulnerabilities in AI systems and making them more robust. His work covers adversarial examples, jailbreaking techniques for LLMs, and understanding the failure modes of models. He has published best papers at IEEE S&P, USENIX Security, and ICML and received extensive coverage by mainstream media platforms including the New York Times, the BBC, Nature Magazine, Science Magazine, Wired, and Popular Science.

For virtual attendees, we will send out the Zoom link 2 hours ahead of the talk.