Bruce Schneier on Why the US Can’t Regulate AI
Watch the Video on YouTube.com
What happens when an autonomous AI agent causes damage without breaking a single security rule?
NVIDIA recently introduced Sentry, a hardware-based security system designed to contain AI agents that attempt to operate outside their authorized boundaries. But as AI agents gain access to sensitive company data, financial systems, and critical infrastructure, is preventing them from escaping their sandbox enough?
In this episode of RegulatingAI, we speak with renowned cybersecurity expert and cryptographer Bruce Schneier about the emerging security challenges of agentic AI.
We discuss the limits of NVIDIA’s approach, whether traditional cybersecurity principles can keep pace with increasingly autonomous systems, and who should be held accountable when AI agents cause harm. We also explore what effective AI regulation should look like and the implementation of the EU AI Act.
As AI moves from generating information to independently taking action, the question is no longer just whether we can secure these systems, but whether our existing security frameworks are sufficient.
Categories: Recorded Interviews, Video