Hacker And The Fed AI Agents Gone Rogue: The Next Breach Waiting to Happen
Aug 28, 2025
The discussion dives into the risks of AI agents potentially causing breaches by clicking on malicious links. The hosts debate the implications of a new bill proposing 'cyber privateers' for hackback actions. They also dissect the DOJ's successful takedown of the RapperBot botnet, highlighting its capabilities. A concerning zero-day flaw affecting many password managers is examined, along with the dangers of hoarding vulnerabilities that can lead to catastrophic leaks. Essential insights into cybersecurity risks and defenses make this an engaging listen!
AI Snips
Chapters
Transcript
Episode notes
AI Agents Can Become Attack Vectors
- Agentic AI deployed as replacement workers can autonomously click malicious links and cause breaches.
- Hector and Chris warn this is already starting and will accelerate with widespread agent use.
Chatting With Grok In A Tesla
- Chris jokingly tested Grok in his Tesla and found it responded with show knowledge about their podcast.
- Both hosts considered inviting Grok for an interview but noted legal/technical hurdles.
Neo-Privateers Are Authorized Hackback
- The proposed neo-privateer bill is effectively authorized hackback with presidential letters of marquee.
- Hector and Chris worry broad scope, attribution and domestic targeting create major legal and ethical risks.