Notes d’épisode
Itay Peled (Project Perception) and Yuri Diogenes (Defender for Cloud) on the two sides of AI in security: using AI to defend, and securing the AI everyone is deploying. Recorded live at Workplace Ninja Summit 2026 with Frans Oudendorp and Alex Verboon.
In this episode
- Why "AI for security" and "security for AI" are different problems that you can't solve separately
- Exposure management for agents: an agent that was never meant to be internet-facing can still be reachable through an attack path
- Prompt injection and indirect prompt injection as the big threat to agents, and why MITRE…
Site web: podcasters.spotify.com

