OpenAI Disrupts Alleged Moonshot AI Model-Distillation Campaign
Emerging Technology Risk · 9 sources · Updated 1mo ago · First seen Jul 28, 2026

OpenAI said it disrupted a coordinated model-distillation campaign in July that it primarily attributed to a cluster associated with Chinese AI developer Moonshot AI, creator of Kimi. The alleged operation used coordinated, high-volume prompts to extract protected model reasoning exposed to requesters, rather than accessing encrypted systems, databases, or user conversations. OpenAI reported approximately 16,000 extraction-pattern requests from more than 4,000 users on July 24–25, with related prompt activity spanning over 15,000 users.
OpenAI said it banned accounts involved in the activity on July 28, strengthened monitoring and controls, and closed a replay pathway involving encrypted reasoning. The company shared investigative details with industry and government information-sharing groups. The reported activity highlights the risk that sophisticated adversaries may seek to replicate proprietary model capabilities through API-driven extraction rather than conventional network intrusion.


