TechCrunch’s Disrupt stage hosts a security panel on the uncomfortable middle of enterprise AI: models are already making autonomous decisions inside sensitive systems, while the controls around them are still being written. Speakers include security leaders from Luta Security and AWS, alongside cybersecurity veteran Wendy Nather.
The session focuses on architecture and governance, meaning what has to be true in the stack before an agent is allowed to act, and who owns the decision when it gets one wrong.
