China is weighing a far-reaching expansion of technology export controls that would put artificial intelligence at the centre of a widening national-security campaign.

The proposal emerged as China has made visible gains in the AI race, with releases such as Kimi K3 underscoring the country’s push to close the gap with US developers. Beijing is reportedly considering placing 23 AI-related technologies on its export-control list, spanning chip designs, AI agents, model weights and source code.¹

The move would mark a notable turn in a contest where China has long been on the receiving end of Western restrictions on advanced AI chips and related tools. From Beijing’s perspective, tighter control over home-grown capabilities is a way to manage security risks — and prevent strategically valuable know-how from flowing abroad as competition sharpens.

On Friday, Chinese authorities outlined an even broader package: plans to restrict or prohibit exports of 40 technologies across AI, semiconductors and solar power. The measures were explicitly tied to national-security concerns and to an intensifying technology race with the United States and other countries.²

Taken together, the reports point to a policy extending beyond hardware. Controls on model weights, source code and AI agents would reach into the software and intellectual-property layers of the sector, where technological advantage can travel faster than a physical chip.

The scope remains unclear, including which items would face an outright ban and which would require licences. But the direction is unmistakable: as Washington seeks to limit China’s access to key technologies, Beijing is preparing to use export controls to protect — and wield — its own.