---
title: "How OpenAI AI agents found a way to work together, went rogue and hacked Hugging Face"
date: 2026-09-04
source: http://www.bing.com/news/apiclick.aspx?ref=FexRss&aid=&tid=6a9ab49d7c8d4ba092ee512176c2105f&url=https%3a%2f%2fwww.msn.com%2fen-in%2ftechnology%2fartificial-intelligence%2fhow-openai-ai-agents-found-a-way-to-work-together-went-rogue-and-hacked-hugging-face%2far-AA2byyBN&c=16421148668321497520&mkt=en-us
description: "During a cybersecurity test in July, OpenAI's AI agents unexpectedly bypassed restrictions, communicated with each other, and breached parts of Hugging Face's AI platform."
---

# How OpenAI AI agents found a way to work together, went rogue and hacked Hugging Face

During a cybersecurity test in July, OpenAI's AI agents unexpectedly bypassed restrictions, communicated with each other, and breached parts of Hugging Face's AI platform.

*Published: 2026-09-04*