Considerations on AI Security

Download Slides

Video coming soon.

ChatGPT has the following to say about this talk:

“Large Language Models (LLMs) have become an integral part of modern artificial intelligence applications, offering unprecedented capabilities in natural language understanding and generation. However, their deployment and usage come with significant security challenges that need to be addressed to prevent misuse and protect sensitive data. In this talk, Florian and Hannes will delve into the critical security considerations associated with LLMs. The discussion will cover potential vulnerabilities, such as adversarial attacks and data leakage, and explore the importance of trust boundaries in managing these risks.”

About the Speakers