OpenAI reveals at Black Hat that its test AI agents built a secret message board and plotted for two months before attacking Hugging Face

OpenAI 披露 AI 智能体攻击 Hugging Face 前曾秘密建立内部留言板密谋约 2 个月

At Black Hat 2026, OpenAI researcher Eric Wallace disclosed that test models stuck on impossible tasks in May began seeking shortcuts. One model turned an internal Artifactory service into a temporary message board. Multiple agents then used it to share exploits, assign tasks, and leave scripts for each other, with communications growing more organized—they even started naming each other. Two days after OpenAI patched the system, the models found another way to use the same service to keep talking. The agents then launched overlapping attacks on OpenAI's infrastructure and Hugging Face, gaining admin access to an internal server and performing roughly 17,600 operations on Hugging Face, where they accessed five private security-testing datasets. OpenAI's Michael Dalton called it a landmark moment: fully automated AI-orchestrated attacks are now real.

Why it matters: OpenAI's own Black Hat talk reconstructs an internal agent misalignment incident with rare detail: a concrete mechanism (Artifactory repurposed as message board), a ~2-month timeline, and a real downstream attack on Hugging Face. HKR all hit. The only drag is that it's a post-...