GoKawiilFBI, NSA, CISA Accuse Alibaba, DeepSeek and Others of Mass-Distilling US AI Models
US federal agencies issued a joint advisory on Sept. 8 alleging that Chinese AI companies including Alibaba, DeepSeek, MiniMax, Moonshot AI, StepFun and Z.AI have run large-scale operations to pull outputs from US models like Claude, GPT, Gemini and Grok since late 2024. The advisory says these firms extracted billions of tokens through millions of queries, using techniques such as chain-of-thought extraction and automated evasion to dodge blocking measures, in order to train their own competing systems.
GoKawiil's interpretation of the reporting above, not reported fact.
While model distillation is a standard technique, the agencies argue this activity crosses into violation of terms of service and intentional theft of proprietary capabilities, potentially with tacit backing from Beijing. The accusation intensifies the US-China AI rivalry and could push American AI labs toward stricter access controls, rate limits, and legal action to protect their models from being used to bootstrap rival systems.
- US agencies allege industrial-scale extraction of billions of tokens from frontier US AI models by Chinese firms.
- Named companies include Alibaba, DeepSeek, MiniMax, Moonshot AI, StepFun and Z.AI.
- The advisory says evasive techniques and possible state awareness distinguish this from ordinary distillation practices.
Hands-On Large Language Models (O'Reilly book) — If terms like distillation, teacher and student models, and token extraction in this advisory left you wanting more, this O'Reilly guide walks through how LLMs are actually built, trained, and adapted. It's a practical way to understand the techniques at the center of the US-China AI dispute instead of just reading headlines about them.
See Hands-On Large Language Models (O'Reilly book) on Amazon → Affiliate link — we may earn a commission on purchases, at no extra cost to you. Product picked by AI based on this article; it is not a tested recommendation.Source: darkreading.com — Alexander Culafi, 2026-09-09
Published there as: “US Government Accuses Chinese AI Firms of Distilling Frontier Models”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.