Beijing is "marshalling data to drive productivity ... (and) improve its intelligence collection ‌and military capabilities," the U.S.-China Economic and Security Review Commission wrote in a report released on Tuesday.

China's National Data Administration did not immediately respond to a faxed request for comment outside of working hours.

While major U.S. AI companies have largely exhausted the open internet for ​language model training, China is systematically collecting domestic "enterprise, operational and physical-world data that cannot be scraped" — crucial ​for training AI tools designed for business use, autonomous vehicles and humanoid robots, a major ⁠symbol of Beijing's top-down innovation push.

The report said for example that China's advanced manufacturing and industrial robotics ecosystems provide ​a vast pool of high-quality data for use in embodied AI applications — AI systems that operate in the physical ​world, such as robots — giving it a potential advantage over the U.S. in developing robotics software for commercial and military uses.

"Over the last half decade, China has been able to consolidate data, find ways to label it and refine it, and hoover up new data ​and make sure it's quickly made available to entities," said Mike Kuiken, the commission's vice chair, in an interview.

"It ​has the benefit of advancing their innovation ecosystem, and it has the benefit of enhancing the control of the (Chinese Communist) Party."

China ‌has tightened ⁠its regulatory oversight of cross-border data transfer in recent years, leading to widespread compliance issues for foreign multinational firms operating in the country.

In order to be in compliance with China’s cross-border data regulations, the firms had to further localise their China subsidiaries and separate their China customer data from their global operations, although Beijing has since introduced limited exemptions for some ​firms.

U.S. firms operating in China "risk ​running afoul of China's ⁠strict data and cyber governance regimes," the report said.

China has elevated data to a core "factor of production" alongside land, labour, capital and technology, and it established the National Data ​Administration body in 2023 that oversees nationwide data standardisation and classification and aims to ​build a unified ⁠national market for data trading.

Beijing has also pursued the commercialisation and deployment of data in its digital economy, alongside a broader push to deploy AI across all industries.

It aims to standardise data assets across critical industries including manufacturing, transport, finance and healthcare, ⁠while ​Chinese firms have begun to list their proprietary datasets on regional data ​exchanges in cities including Shanghai, Shenzhen and Beijing.

"We recommend that U.S. Congress think about a national data strategy and how the U.S. government could ​treat data as an economic asset as our number one recommendation," Kuiken said.