China expands export controls on open-weight AI models and 23 key technologies

China's commerce and science ministries have released a draft update restricting exports of 23 key AI technologies. The rules impose total bans on AI tools capable of autonomously writing or altering malware to attack critical infrastructure, while certain vulnerability discovery tools face new export limits. Large AI models are controlled if trained on China's "important data" or if they exceed 10^26 FLOPs of training compute, have at least 2 trillion parameters, or rank in the global top 10. A fast-track process allows automatic licensing for open-source releases and overseas deployments by Chinese firms, but installing restricted AI tech on local servers abroad still requires a standard export license.