Hermes Vector

AI & SI

Chinese developers use US AI models to train domestic rivals as chip curbs tighten

A Reuters review of more than 300 documents shows Chinese companies using output from leading US AI models to train domestic alternatives through distillation, as Beijing seeks cheaper AI systems amid tighter US chip restrictions.

By Hermes Vector

Also in: Español · Português · Français · Deutsch

Checked against 2 sourceseconomictimes.indiatimes.comyahoo.com
Illustration for: Chinese developers use US AI models to train domestic rivals as chip curbs tighten

In brief

  • Chinese companies use output from US AI models to train domestic alternatives through distillation.
  • A Reuters review of more than 300 documents found the practice has moved into mainstream industrial use in China.
  • Washington believes industrial-scale distillation is the main reason Chinese AI models have narrowed the gap with US rivals.
  • US Commerce Department officials warned in July that some output collection could amount to intellectual-property theft.

Chinese developers use US AI models to train domestic rivals as chip curbs tighten

Chinese entities are using output from leading American AI models to train domestic rivals, according to a Reuters review of more than 300 Chinese academic papers, patents, corporate disclosures, recruitment notices and government documents. The practice, known as distillation, has expanded beyond academic experiments into mainstream industrial use in China as Beijing seeks cheaper AI systems and faces tighter US restrictions on advanced chips.

How distillation works

Distillation involves a powerful "teacher" model generating answers or demonstrating how to complete tasks, which developers use to train a smaller "student" system. For Chinese developers facing constraints on computing power, the attraction is in making one expensive query do repeated work, according to a former Alibaba Group engineer who spoke on condition of anonymity.

The documents describe Chinese researchers using US models to produce answers to questions, build computer code and find solutions to difficult problems, with Chinese companies then developing software to automate the process and specialist workers preparing material for training smaller domestic models.

Specific examples cited in the documents include:

  • In a November 2024 paper, Shanghai Jiao Tong University researchers selected about 5,000 answers from OpenAI's o1 model, translated them using GPT-4o mini and assembled 10,750 bilingual training examples to further train Alibaba's Qwen2.5-72B model.
  • In an April 2025 paper, Alibaba researchers used OpenAI's GPT-4 and GPT-4o to generate training material in languages other than Chinese, then refined and scored responses before training smaller Qwen models.
  • In a February 2026 paper, researchers from Shanghai Jiao Tong University and Chinese food-delivery company Meituan described using GPT-4o to generate plans for coordinating AI agents working together on programming problems, helping train a smaller model to organise such tasks.

Geopolitical dispute

Distillation is not illegal and is prevalent in other parts of the world. While the process cannot reproduce a leading model's full capabilities or eliminate the need for advanced computing power, it can help developers preserve selected skills in smaller systems that cost less to operate, potentially reducing their reliance on American technology.

Washington said on Monday it would further restrict China's access to such chips from companies like Nvidia. US authorities publicly accused Chinese companies in September of aggressive, targeted distillation of American AI models on an industrial scale.

At meetings in Chengdu in July, US Commerce Department officials raised concerns with Chinese regulators about developers collecting output from leading American AI models, warning that some activities could amount to intellectual-property theft, three people familiar with the discussions said. Chinese officials defended distillation as a standard AI-development technique and said they were open to acting where the conduct constituted IP theft, but made no concrete commitments, according to a US official familiar with the previously unreported exchange.

Among leading American AI companies, OpenAI's terms restrict using its output to develop competing models, subject to exceptions, while Anthropic similarly restricts training competing systems without permission.

A US official familiar with the July Chengdu regulatory discussions said Washington believes industrial-scale distillation is the main reason Chinese AI models have narrowed the gap with their American rivals. Two people familiar with Chinese developers' operations disputed that assessment, saying distillation was only one factor alongside advances in engineering, algorithms, data and model design.

Beijing's policy support

Chinese government documents do not instruct companies to distil American models but encourage distillation as part of Beijing's push to spread AI across the world's second-largest economy. A national AI+Manufacturing policy package published by eight Chinese government agencies in January includes guidance supporting model training, fine-tuning and distillation for industrial applications. Shanghai and Shenzhen have also encouraged techniques to reduce computing requirements.

The policies reflect Beijing's ambition to deploy smaller, cheaper AI systems across industries. Concerned by distillation's growing geopolitical sensitivities, Chinese AI developers have in recent months asked Beijing's industry regulator for guidance on the boundaries between ordinary use of foreign models and organised collection of output for training, according to one person familiar with the discussions.

The US Commerce Department and China's Cyberspace Administration did not respond to requests for comment. Shanghai Jiao Tong University, Alibaba and Meituan did not respond to requests for comment.

Why it matters

China's broadening use of distillation could undercut the effectiveness of US curbs on the sale of advanced chips to China, as its developers can build AI capabilities by drawing on those of their American rivals. The issue is a source of friction between the world's two biggest economies ahead of the November 18-19 Asia-Pacific Economic Cooperation summit in Shenzhen, where US President Donald Trump and his Chinese counterpart Xi Jinping are expected to meet.


Sources

This article was drafted with AI assistance and checked against the sources above. Company claims are reported as claims. Cover image is AI-generated.

The day's checked news, by email

One short email a day with the top stories. Free, and you can unsubscribe in one click.

Or follow for the latest news:

How do you feel about this story?

Comments 0

No comments yet. Start the discussion.