OpenAI Attributes Summer Distillation Campaign to Moonshot AI Associates

OpenAI disclosed it detected and disrupted a coordinated campaign to extract protected reasoning from its models, attributing a core cluster of the activity to individuals associated with Moonshot AI, the Beijing-based developer of Kimi. The attack began in the first week of July, peaked at 16,000 requests from more than 4,000 users on July 24-25, and was fully shut down by July 28. Attackers did not break encryption or access databases; instead they copied encrypted reasoning from one conversation and asked the model in a separate conversation to transcribe it, reproducing protected content in readable form. OpenAI deployed additional mitigations, banned the fraudulent accounts, and shared findings with industry partners through the Frontier Model Forum and with government information-sharing channels. Moonshot AI had not issued any public statement as of early October. OpenAI explicitly framed adversarial distillation as a national security concern, arguing that capability transfer without matching safety investment is especially dangerous in dual-use domains. The 16,000-request incident is smaller in scale than distillation attempts reported by Anthropic, which logged incidents in the millions of exchanges. OpenAI and Microsoft previously raised similar concerns about DeepSeek in early 2025. China's government has previously rejected distillation accusations from U.S. companies and the Trump administration and has threatened retaliation against U.S. penalties.