OpenAI Blocks 15,000-User Campaign Trying to Extract Hidden AI Reasoning
OpenAI stopped a 15,000-user campaign, partly tied to Moonshot AI, extracting hidden reasoning.
OpenAI said it disrupted a coordinated effort by more than 15,000 users to extract protected model reasoning through adversarial prompts and cross-conversation replay, not by breaching encryption, databases, or stored chats. Activity began at low volume on July 1 and spiked on July 24–25, when 16,000 matching requests came from more than 4,000 users; OpenAI said the operation was fully disrupted by July 28. It attributed a core cluster to people associated with Moonshot AI, developer of the Kimi model, while saying it could not tie every account to one actor. OpenAI restricted accounts, blocked a reasoning-replay path, and shared information through the Frontier Model Forum and government channels.
- Operators tried to expose hidden reasoning through coordinated prompts, not a database breach.
- OpenAI saw 16,000 matching requests from over 4,000 users on July 24–25.
- The wider cluster exceeded 15,000 users and was disrupted by July 28.
- A core portion was attributed to associates of Moonshot AI, maker of Kimi.
- OpenAI closed a replay path and shared details with industry and government partners.
- Organizations
- OpenAIMoonshot AIFrontier Model Forum
- AI models
- Kimi
From the original article
2 min read at cybersecuritynews.com
OpenAI disrupted a coordinated campaign involving more than 15,000 users that attempted to extract protected reasoning from its AI models through adversarial distillation. Protected reasoning refers to the internal process a model uses to work through a task before producing its final answer. This hidden reasoning may include intermediate analysis, planning steps, and information intentionally […]
By Abinaya · cybersecuritynews.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.