AGI Dreams Podcast – July 28, 2026

AGI Dreams – Open, Uncensored, & Local - AI News Digest · Jul 28, 2026 · 17m · transcript

AI Security: Autonomous Ransomware and the Agentic Red Team

In this episode:

AI Security: Autonomous Ransomware and the Agentic Red Team — The first documented end-to-end autonomous LLM ransomware operation exploited a Langflow RCE flaw, while defenders like Lovable deploy AI hacking agent swarms against their own infrastructure. New offensive toolkits and debates over frontier-model security guardrails intensify the offense-defense duality.

Governing AI with Temporal Logic — University of Toronto researchers introduce TRAC, which uses Linear Temporal Logic to formally monitor…