Anthropic details distillation campaigns from Alibaba, Moonshot AI, and DeepSeek
Anthropic reports nearly 200 million Claude exchanges tied to distillation campaigns by Alibaba, Moonshot AI, and DeepSeek.
A new Anthropic report describes five distillation campaigns totaling nearly 200 million exchanges that extracted chain-of-thought traces from Claude to train competing models, targeting agentic tool use, coding, and reasoning capabilities. The largest campaign, attributed to Alibaba, accounted for 151 million exchanges between May and July 2026 across 3,500 accounts, peaking near three million exchanges per day, allegedly to produce training material for the Qwen model family. A Moonshot AI campaign routed roughly 300,000 requests over ten days through 5,000 accounts, primarily targeting Opus, including one task analyzing CCTV footage that appeared connected to the Chinese military. Attackers used prompt techniques, such as framing queries as katakana-only Japanese translation requests, to make Claude reveal its internal thinking traces.
- Anthropic attributes nearly 200 million Claude exchanges across five campaigns to unauthorized model distillation
- Largest campaign, attributed to Alibaba, ran 151 million exchanges May-July 2026 across 3,500 accounts
- Moonshot AI campaign routed ~300,000 requests through 5,000 accounts over ten days, one assessing CCTV footage
- Attackers used prompt tricks, like katakana-only Japanese translation requests, to expose Claude's chain of thought
- Campaigns targeted agentic tool use, coding, data analysis, and logical reasoning capabilities
- Organizations
- AnthropicAlibabaMoonshot AIDeepSeek
- Countries
- China
From the original article
2 min read at techcrunch.com
A new report released Thursday by Anthropic alleges persistent distillation attacks by China-based AI companies, which have escalated in recent months as competition in the space has intensified.
By Russell Brandom · techcrunch.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.