Nobody Signed Up to Govern A Robot

Name: Nobody Signed Up to Govern A Robot
Start: 2026-09-30T13:30:00-05:00
End: 2026-09-30T14:00:00-05:00
Location: Texas Public Radio

When: Wed, Sep 30 · 1:30 – 2 PM
Where: Texas Public Radio

Most data risk programs were built for a world of static systems, predictable data flows, and controls that didn’t move once you set them. AI broke that assumption. Models retrain, infer, and act in ways that outpace the policies written to contain them and the risk function has had to catch up in real time. This session walks through what’s actually changed in data security, AI governance, and AI security since AI adoption accelerated across the enterprise. Not the theoretical risks, but the practical shifts: how data classification has to become an active trigger instead of a static label, how governance checkpoints need to move with the AI lifecycle (design, build, deploy, monitor, retire) instead of sitting at the end of it, and how new attack surfaces, prompt injection, model manipulation, data leakage through inference, demand a different kind of vigilance than traditional infosec controls were built for. Attendees will leave with a practical view of how to evolve an existing data risk program to meet AI where it is, rather than trying to bolt AI governance on as an afterthought, grounded in real program-building experience, not just frameworks on paper.

Tech & Builders circuit, sponsored by

The rest of the day Everything at TPR