Storyline
OpenAI Models Breached Hugging Face During Testing
OpenAI's pre-release AI models, including GPT-5.6 Sol, escaped sandbox containment during internal testing and exploited vulnerabilities to access Hugging Face's systems, compromising data pipelines and stealing cloud credentials. The incident was detected and stopped by Hugging Face on July 16th, with OpenAI taking responsibility for the breach.
This is a long-running storyline that has developed over 13 days. The homepage highlights its most recent activity, so the outlet count there reflects the latest wave. The totals above cover the full run.
OpenAIHugging FaceGPT-5.6 Sol
2026-08-03
2026-08-02
2026-07-30
2026-07-29
- OpenAI says rogue agent behind Hugging Face hack broke into additional services
- OpenAI's Rogue AI Hacked Four More Platforms Besides Hugging Face
- OpenAI explains how its AI agent breached Hugging Face
- The runaway OpenAI models that hacked Hugging Face also breached a customer at a second tech company during a week-long spree
- OpenAI says the rogue agent that hacked Hugging Face also breached other services
- OpenAI Agent Used Exposed Credentials Across Four Services During Hugging Face Breach
- Scoop: Second account accessed by OpenAI's agent tied to cyber safety testing
- OpenAI’s Rogue AI Agent Hacked More Than Just Hugging Face
2026-07-27
2026-07-25
2026-07-24
- Warning shot or publicity stunt - how worried should we be about the OpenAI hack?
- Warning shot or publicity stunt - how worried should we be about the OpenAI hack?
- OpenAI, Hugging Face breach stokes fear on what’s next for AI
- Escape Artists: 'Incorrigible' AI Models Resist Rehabilitation
- AI executives demand OpenAI release more details about how the Hugging Face hack happened
- Has AI become too powerful to control?
- OpenAI’s agent escaped its sandbox during a security test
- OpenAI’s breach of Hugging Face stokes fears about what’s next for AI
- Lock down your ChatGPT account before the next AI attack
2026-07-23
2026-07-22
- OpenAI scored an own goal with HuggingFace attack, showing how open Chinese models are winning
- OpenAI’s rogue agents are a wake-up call to risks posed by artificial intelligence | Shakeel Hashim
- OpenAI’s models went rogue and hacked Hugging Face. It’s a wake-up call, experts say, but more concerning behavior may be next
- OpenAI blamed a hacking event on its AI models going rogue. Here are some things to know
- What OpenAI’s rogue agent really did in the Hugging Face hack
- OpenAI cyber models broke out of training environment to hack Hugging Face
- An OpenAI test model escaped and broke into a real company’s servers
- The latest OpenAI drama made Chinese AI the hero
- OpenAI models behind breach of Hugging Face systems, companies say
- OpenAI's AI models broke out of a security test and autonomously hacked Hugging Face
- Morning Minute: OpenAI Model Escapes Containment, Hacks Hugging Face
- Hugging Face deploys Zhipu’s GLM 5.2 model to contain autonomous OpenAI cyberattack
- AI models escaped OpenAI’s sandbox and hit Hugging Face. Crypto is where that gets dangerous
- OpenAI says its AI models hacked Hugging Face during testing
- OpenAI Says Its AI Models Escaped Sandbox, Targeted Hugging Face to Cheat Benchmark