Smashing Security podcast #472: AI gets hacked, and BitLocker gets bypassed

This Smashing Security episode reviews multiple high‑risk security issues: Tenet's research showing attackers can plant malicious Sentry error reports that agentic AI coding assistants will treat as trusted input and execute (enabling credential and data theft), and Nightmare Eclipse's public disclosure of zero‑day issues — notably a BitLocker recovery/TPM bypass (Yellow Key) and Microsoft Defender flaws — plus discussion of vendor responses, real‑world testing, and defensive recommendations for organisations integrating AI agents.