Enterprise Audit Finds 9 AI Agents Running
tradami · reddit · 2026-07-15
The author shared that ahead of a company audit in 90 days, they tried to map out how many AI agents were running in their production environment. They discovered that on top of the 4 known agents, 9 agents were actually running. Some of these were secretly deployed by different teams without the engineering team's approval.\n\nWorse, two unregistered agents had write access to the same Postgres instance storing customer PII. The author's concern is that if a record was altered last Tuesday, it would be extremely difficult to quickly trace which agent made the change. The post ends by asking if other teams have faced similar issues during their audits.
Related event: Enterprise AI agent audits expose shadow governance gaps(5 posts)→
More from coding & agent
- Agent verified 'success' after double-voting: a lesson in harness vs model bugs — TejasKumar_ · 2026-10-26
- diginomica: agentic automation hype is overdone, and 'The Hackening' of enterprise security has begun — jonerp · 2026-10-26
- 'Vibe verification': ~60% of AI-written formal specs prove the wrong property — basedjensen · 2026-10-26
- Devin launches Security suite powered by Claude Mythos with $100M+ credits — DevinAI · 2026-10-26
- HackBench injects reward hacks into agent trajectories to stress-test monitors — DanielKhashabi · 2026-10-26
- Hugging Face ships hf CLI updates: deduped cache saves 19GB, faster buckets, retryable GPU jobs — yb2698 · 2026-10-26