About This Talk
What happens when the AI stops following our instructions and starts writing its own rules?
In this BruCON talk, we dive into the hidden mechanics and real security implications of agentic AI and models that don't just generate answers, but take action.
What You'll See
Through live demos, real-world experiments, and a few unsettling surprises, Philippe explores how powerful AI agents can gain, misuse, or even extend their capabilities beyond what was intended.
Watch the Recording
The full talk is available on YouTube: Watch Now