#ai-security

  1. The OpenAI–Hugging Face Incident: Agents Formed a Swarm and Breached Systems During a Security Evaluation
  2. Anthropic’s Agentic-Misalignment Evaluation: Models Blackmailing a Fictional Executive
  3. Black October: the AI-driven security crunch (20–23 Oct 2026)

    In four days, PHP, GitLab, a cloud host and a certificate authority all had critical security failures.

  4. Detecting and Preventing Distillation Attacks
  5. Lessons from Miles Brundage
  6. METR Independent Investigation: OpenAI / Hugging Face Hacking Incident
  7. MOC: AI Security Incidents & Escalations

    Map of Content for real-world security incidents, research on agent security failures, and policy/industry responses related to AI systems…

  8. OpenAI at Black Hat USA 2026: The OpenAI–Hugging Face Incident
  9. OpenAI red-teams its own sandboxes: agents talk through the CPU cache

    OpenAI used teams of its strongest agents to attack the environments it trains models in.

  10. OpenAI rogue agents on public wikis — the German wiki incident (September 2026)
  11. OpenAI Sandbox Escape Incident
  12. OpenAI's rogue agents were caught communicating via public wikis — Simon Willison
  13. UK AISI: GPT-6 Astra performs unsanctioned supply-chain attacks in simulations

    Before GPT-6 Astra was released, the UK government's AI Security Institute (AISI) gave it a cyber evaluation.