{"version":1,"type":"story","url":"https://digestai.news/story/anthropic-reports-distillation-attacks-by-chinese-ai-labs","json":"https://digestai.news/story/anthropic-reports-distillation-attacks-by-chinese-ai-labs.json","markdown":"https://digestai.news/story/anthropic-reports-distillation-attacks-by-chinese-ai-labs.md","slug":"anthropic-reports-distillation-attacks-by-chinese-ai-labs","headline":"Anthropic reveals industrial-scale Claude distillation attacks by seven Chinese AI labs","summary":"Anthropic has disclosed that seven China-based AI laboratories, including Alibaba, DeepSeek, and Moonshot, conducted large-scale illicit distillation campaigns against its Claude models. These operations, detected since February 2026, involved the unauthorized extraction of model capabilities, specifically chain-of-thought reasoning and agentic skills, to train rival systems. The attacks utilized sophisticated evasion tactics, including networks of thousands of fraudulent accounts, stolen API keys, and proxy services that rerouted user traffic without consent.\n\nThe most significant incident, labeled GTG-16005, involved Alibaba-affiliated operators harvesting 151 million exchanges between May and July 2026, peaking at 3 million daily requests. Other campaigns included Moonshot and DeepSeek silently relaying customer queries to Claude to capture training data, while Z.ai and Xiaomi employed similar replay and extraction pipelines. Anthropic noted that these labs also purchased user transcripts from third-party resellers who harvested data from proxy networks.\n\nIn response, Anthropic has implemented technical defenses, including updating Claude to summarize internal reasoning before responding to reduce the utility of stolen transcripts. The company also introduced \"preserved thinking\" in Fable 5.1 to prevent prompt manipulation. This disclosure aligns with recent accusations from U.S. intelligence agencies regarding systematic extraction of proprietary AI capabilities by Chinese firms.","keyPoints":["Anthropic identified six illicit distillation campaigns by seven Chinese labs, including Alibaba, DeepSeek, and Moonshot, since February 2026.","The largest attack, GTG-16005, involved 151 million exchanges and 3,500 fraudulent accounts targeting Claude Opus 4.6 and 4.7 reasoning.","Labs used proxy services, stolen credentials, and silent traffic rerouting to harvest chain-of-thought data for unauthorized model training."],"whyItMatters":"This exposes a critical security vulnerability in frontier AI, where proprietary reasoning capabilities are being systematically stolen to accelerate competitor development. It highlights the rising sophistication of industrial-scale AI espionage and the urgent need for robust technical and legal defenses against model extraction.","category":{"slug":"models","name":"Generative AI & Models","url":"https://digestai.news/category/models"},"entities":{"companies":["Anthropic","Alibaba","DeepSeek","Moonshot","Z.ai","MiniMax"],"models":["Claude","Claude Opus 4.6","Claude Opus 4.7","Fable 5.1"],"people":[]},"firstPublishedAt":"2026-09-10T20:57:30Z","updatedAt":"2026-09-11T16:59:00Z","sourceCount":9,"hasPrimarySource":false,"sources":[{"outlet":"yourstory.com","title":"China’s AI labs siphoned Claude’s intelligence?","url":"https://yourstory.com/2026/09/chinas-ai-labs-siphoned-claudes-intelligence","publishedAt":"2026-09-11T16:59:00Z","type":"press","primary":false,"lead":true},{"outlet":"ibtimes.com","title":"DeepSeek And Alibaba Are Closing The AI Gap. Anthropic Accuses them Of Using Claude To Help Train Their Models.","url":"https://ibtimes.com/deepseek-alibaba-are-closing-ai-gap-anthropic-accuses-them-using-claude-help-train-their-3807381","publishedAt":"2026-09-11T10:17:00Z","type":"press","primary":false,"lead":false},{"outlet":"cryptobriefing.com","title":"Anthropic disrupts massive distillation attack from Chinese AI labs targeting Claude","url":"https://cryptobriefing.com/anthropic-disrupts-chinese-ai-distillation-claude","publishedAt":"2026-09-11T09:27:00Z","type":"press","primary":false,"lead":false},{"outlet":"thehackernews.com","title":"Anthropic Says Seven China-Based AI Labs Ran Industrial-Scale Claude Distillation Attacks","url":"https://thehackernews.com/2026/09/anthropic-says-seven-china-based-ai.html","publishedAt":"2026-09-11T09:16:00Z","type":"press","primary":false,"lead":false},{"outlet":"The Rundown AI","title":"Anthropic opens the files on global Claude misuse","url":"https://therundown.ai/articles/anthropic-opens-the-files-on-global-claude-misuse","publishedAt":"2026-09-11T09:00:00Z","type":"newsletter","primary":false,"lead":false},{"outlet":"CNBC Technology","title":"Chinese AI labs secretly used millions of Claude exchanges to train their models, Anthropic says","url":"https://cnbc.com/2026/09/11/chinese-ai-labs-moonshot-deepseek-alibaba-anthropic.html","publishedAt":"2026-09-11T00:59:45Z","type":"press","primary":false,"lead":false},{"outlet":"businessinsider.com","title":"China's star AI labs routed user requests to Claude at least 35 million times in the summer: Anthropic","url":"https://businessinsider.com/china-ai-labs-millions-distillation-attacks-anthropic-claude-2026-9","publishedAt":"2026-09-10T23:48:00Z","type":"press","primary":false,"lead":false},{"outlet":"TechCrunch AI","title":"Anthropic details distillation campaigns from Alibaba, Moonshot AI, and DeepSeek","url":"https://techcrunch.com/2026/09/10/anthropic-details-distillation-campaigns-from-alibaba-moonshot-ai-and-deepseek","publishedAt":"2026-09-10T20:57:30Z","type":"press","primary":false,"lead":false},{"outlet":"nypost.com","title":"Anthropic caught Chinese AI labs carrying out massive ‘distillation attack’ to rip off its technology","url":"https://nypost.com/2026/09/11/business/anthropic-caught-chinese-ai-labs-carrying-out-massive-illicit-distillation-attack","publishedAt":"2026-09-10T20:20:00Z","type":"press","primary":false,"lead":false}],"sourceNotes":null,"discussions":[{"site":"Hacker News","url":"https://news.ycombinator.com/item?id=49650135","points":6}],"thread":{"title":"AI Arms Race Theft Misuse Open‑Source Scrutiny","url":"https://digestai.news/thread/us-agencies-accuse-six-chinese-ai-firms-of-largescale-model-theft","storyCount":10},"cite":{"text":"Digest AI, \"Anthropic reveals industrial-scale Claude distillation attacks by seven Chinese AI labs\", 10 September 2026, https://digestai.news/story/anthropic-reports-distillation-attacks-by-chinese-ai-labs","publisher":"Digest AI","title":"Anthropic reveals industrial-scale Claude distillation attacks by seven Chinese AI labs","datePublished":"2026-09-10T20:57:30Z","url":"https://digestai.news/story/anthropic-reports-distillation-attacks-by-chinese-ai-labs"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}