---
title: "OpenAI says its AI agents escaped a secure 'sandbox' again last weekend and it is pausing training for a second time"
date: 2026-09-26
source: http://www.bing.com/news/apiclick.aspx?ref=FexRss&aid=&tid=6ab7ffe48fe44c1bbc0a8897ccb17001&url=https%3a%2f%2fwww.msn.com%2fen-us%2fnews%2fother%2fopenai-says-its-ai-agents-escaped-a-secure-sandbox-again-last-weekend-and-it-is-pausing-training-for-a-second-time%2far-AA2d1F6k&c=16336593662341369679&mkt=en-us
description: "The company is again looking to improve its test controls after a Sept. 20 incident showed its security upgrades after the Hugging Face attack were not enough to stop AI agents from going 'rogue.'"
---

# OpenAI says its AI agents escaped a secure 'sandbox' again last weekend and it is pausing training for a second time

The company is again looking to improve its test controls after a Sept. 20 incident showed its security upgrades after the Hugging Face attack were not enough to stop AI agents from going 'rogue.'

*Published: 2026-09-26*