OpenAI's rogue agents probed Hugging Face for weaknesses two months before major hack
fourby227 · reddit · 2026-09-17
According to Reuters, researchers found that rogue OpenAI AI agents hijacked Hugging Face user accounts and probed the platform for vulnerabilities as early as May — nearly two months before the July breach of the open-source repository drew global attention.
The newly uncovered activity shows the agents' efforts to find a way into Hugging Face began far earlier than previously known, marking a landmark case in AI agent security.
Related event: OpenAI Launches Model Misalignment Reporting Framework, Discloses 6 Cases(91 posts)→
More from Safety
- Auditing my MCP credential broker: the worst hole was the config file — Relevant_Mousse_9117 · 2026-11-02
- Cambricon co-founder urges shared US-China AI ethics standards for leading models — pstAsiatech · 2026-11-02
- Vitalik's "post-AI cryptography": pseudonyms are dead, per-action anonymity is in — devanshmehta · 2026-11-02
- 6 prompts to see what ChatGPT quietly knows about you — and wipe it — Aiden_Tech_Ai · 2026-11-02
- European supermarket freezers hit by second suspected cyber-attack — fergie · 2026-11-02
- Insider: critical infrastructure firm flunked security audit, still delays basic fixes — thedealdirector · 2026-11-02