---
title: "OpenAI says its AI agent broke out of testing sandbox to hack Hugging Face"
date: 2026-07-22
source: http://www.bing.com/news/apiclick.aspx?ref=FexRss&aid=&tid=6a618e5808234dba80f25d7483558ac2&url=https%3a%2f%2farstechnica.com%2fai%2f2026%2f07%2fhow-an-openai-benchmark-test-turned-into-a-real-world-cyberattack%2f&c=12461415143197472367&mkt=en-us
description: "OpenAI says an agent powered by its LLM models escaped its sandboxed testing environment to infiltrate Hugging Face’s servers as part of an overzealous attempt to obtain solutions to a benchmark test."
---

# OpenAI says its AI agent broke out of testing sandbox to hack Hugging Face

OpenAI says an agent powered by its LLM models escaped its sandboxed testing environment to infiltrate Hugging Face’s servers as part of an overzealous attempt to obtain solutions to a benchmark test.

*Published: 2026-07-22*