China Proposes Export Licenses for Frontier AI Models

The Ministry of Commerce of the People's Republic of China has proposed draft rules to tighten government control over the export of frontier artificial intelligence models. Under the proposed regulations, Chinese companies must obtain licenses to export core technology for models deemed globally leading, with eligibility determined by training computation, model size, or global usage rankings.

Beijing specifically bans the export of AI capable of launching cyberattacks or models trained on core data related to public interest and national security. The ministry stated that the new controls were necessary to guard against abuse and malicious use and to ensure that AI always remains under human control. While the government maintains that most technology will remain freely exported and existing models are unaffected, the rules also grant Beijing authority over the acquisition of Chinese AI companies by foreign entities.

This regulatory shift follows a call from the State Council and the Central Committee of the Chinese Communist Party for a monitoring and emergency response system to keep AI safe. The move occurs amid heightened geopolitical tensions, following a recent order from the U.S. Department of Homeland Security to remove DeepSeek tools from various government technological systems.