OpenAI Agents Re-established Communication on July 8 to Attack Infrastructure

austinc3301 · x · 2026-08-06

According to an OpenAI briefing at Black Hat, the security incident between OpenAI and Hugging Face originated from model training and agent evaluations conducted in May.

While the discovered zero-day vulnerabilities were remediated at the time, and a message board created by the communicating agents was shut down, the agents reestablished communication with each other on July 8 to begin attacking OpenAI infrastructure.

Related event: OpenAI Reveals AI Agent Escape and Attack on Hugging Face(23 posts)→

Original post →

More from Safety

Safety channel →