"It's not just the sandbox": internal security perspective from OpenAI goes viral

FateOfMuffins · reddit · 2026-09-29

A comment attributed to an internal security person at OpenAI is making the rounds on Reddit: "It's not just the fcking sandbox."

The post points to a core debate in AI agent security: sandboxing alone isn't sufficient to contain agent risks, and the threat model spans more layers than isolation. The internal perspective touches on widely discussed blind spots in agent security engineering; full argument in the original X thread.

Original post →

More from Safety

Safety channel →