Anthropic says DeepSeek, Xiaomi, and Moonshot used Claude outputs for model distillation

Xi going to India; Plan for Building a Financial Powerhouse; Anthropic drops bombs in PRC AI firms; “Enemy-state clauses”; US-China

Anthropic's September threat-intel report calls out DeepSeek, Xiaomi, and Moonshot for piping user-model conversations into Claude and using Claude's replies as training data for distillation. The exchanges reportedly contained sensitive info from individual users, multinationals, and state-affiliated actors. Anthropic says this violates PRC law and suggests sharing detailed findings with China's Ministry of Public Security via the FBI. The post doesn't disclose the volume of conversations or the time range involved.

Why it matters: Anthropic's official threat intel report names three major Chinese AI labs for distilling Claude with sensitive user data — an industry-level security incident. Strong cross-source signal, all three HKR axes hit. The slight deduction is because we only have Sinocism's second-h...