Vibeleaderboard
← All Intel
Intel / podcast

LIVE From CrowdStrike's Fal.Con 2026 | George Kurtz, Michael Sentonas, Daniel Bernard & Justin Boitano

Source
share.transistor.fm
Author
John Coogan & Jordi Hays
Date

Context

CrowdStrike CEO George Kurtz put visibility into an 's actions at the center of AI-agent security in the opening interview of the TBPN show's broadcast from CrowdStrike's Fal.Con 2026 conference. Kurtz said the company can record every action an agent takes with a specific identity: every tool call, every spawned agent, every network connection and every prompt. That would give defenders a record of what an automated worker actually did.

The interview also distinguishes analyzing activity from stopping it. Kurtz said models are good at sorting out what happened and sifting through data, but that deterministic security must sit in line with the activity and make a decision the first time. Kurtz said that, to Kurtz's knowledge, no model had stopped a breach as it happened. These are the company's claims in a conference interview, not tested coverage guarantees.

Terms in this piece · Glossary
  • AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
Read the source share.transistor.fm
More from John Coogan & Jordi Hays
Recommended reads
Comments

Checking sign-in…

Loading comments…