Episode
0075: An AI System Was Used To Breach 30 Companies And Here’s What It Means For Your Security
- Podcast
- #TheTechHustle Podcast 🎙
- Published
- Nov 27, 2025
- Duration seconds
- 3581
- Processing state
not_requested
Actions
POST https://stenobird.com/v1/public/podcasts/thetechhustle-podcast-6701182/episodes/0075-an-ai-system-was-used-to-breach-30-companies-and-here-s-what-it-means-for-your-security/transcription-requests
Idempotently request low-priority transcript generation for this episode.GET https://stenobird.com/podcast/thetechhustle-podcast-6701182/0075-an-ai-system-was-used-to-breach-30-companies-and-here-s-what-it-means-for-your-security.md
Read the agent-friendly Markdown representation of this episode resource.
Summary
An AI lab just caught something many security teams feared: coordinated agents using jailbreak prompts to bypass safety rules and help breach real systems. We unpack how Anthropic spotted the activity, why multi-agent workflows compress months of human recon into hours, and what that speed advantage means for anyone running internet-facing infrastructure. We take you through the modern attack chain from the defender’s view—asset discovery, version fingerprinting, CVE mapping, exploit testing...