Summit Keynote: Exfil Everything: A Year of Stealing Data from AI Agents

Your AI agent has the same trust boundary problem your browser had in 2005. Are you sure it’s fixed?

Join Ads Dawson of Dreadnode and Mike Takahashi of Zenity — ranked among the top bug bounty hunters on HackerOne and Mozilla’s AI programs — as they break down a year’s worth of data exfiltration bugs found across production AI applications, showing how the same primitive behind XSS now lets attackers steal PII, business data, and private messages through AI agents.

In under an hour, you’ll see real attack chains from bug bounty reports, learn bypass techniques for every common mitigation, and walk away with a repeatable hunting methodology you can point at any AI target with a bug bounty program.

If you’ve ever exploited a trust boundary in a browser, you already have the instincts. This talk shows you where to point them next.

Chat with your fellow AI Summit attendees in the Antisyphon Discord server: https://discord.gg/antisyphon

---

 This talk is part of Antisyphon Training’s Infosec: Age of AI Summit, a free, six-hour, live virtual event designed to give you a practical, real-world look at how cyber experts are using AI and what security professionals need to know as adoption grows.

For those who want to go further, high-quality, affordable hands-on workshops are available Aug. 17-18 to help you learn AI security concepts, techniques, and tactics.

See the entire Antisyphon Training Course Catalog for affordable cybersecurity training! 

---