In this episode Jb and Izar are joined by David Haber, CEO of Lakera, who focuses on securing LLMs and their use. We explore topics like prompt injection and their impact on security, safety and trust, and we look at the Gandalf experiment ran by Lakera. We touch on the recently drafted OWASP Top 10 on LLM project, and have a great discussion on what LLMs are really doing and their potential as tools and targets.