Today in AI: Defensive AI Agents Pwned, JadePuffer Ransomware Goes Fully Agentic, Meta Leaks Employee Keystrokes — July 9, 2026
July 9, 2026 — curated links and takeaways. 1. Defensive AI Agents Can Be Tricked Into Running Malicious Code AI Now Institute researchers demonstrated that security-focused AI agents designed to catch malicious …