IT Home October 23 news: On October 16, the Ministry of Commerce and the Ministry of Science and Technology solicited public comment onrevising the Catalogue of Technologies Prohibited or Restricted from Export of China。
Among these, the AI field includeslarge models, agents, computing clusters, and information processing—23 related technologies; technologies that have already entered the public domain through open-sourcing, publication of academic papers, and other means are not included in the scope of management.
Yuyuan Tantian, a new media account under China Media Group, revealed in a post last night (the 22nd) thatsince the policy was announced, the two leading companies DeepSeek and MiniMax have smoothly completed registration procedures and finished open-sourcing and overseas licensing use。
The spokesperson of the Ministry of Commerce previously answered reporters' questions on related matters; IT Home has compiled the original text of the Ministry of Commerce as follows:
Q: What are the main considerations for the AI technologies proposed to be added?
A: At present, global AI technological innovation has entered an unprecedented period of activity, with both opportunities and challenges. The proposed addition of 23 key technologies related to AI large models, agents, computing clusters, and information processing to the Catalogue is mainly based on the following considerations:
First, to ensure AI technologies are safe and controllable. The various endogenous and derivative risks caused by AI have attracted high attention from countries around the world. The AI technologies proposed for control this time, such as AI autonomously driven malicious code generation and mutation technology, could be used to attack sensitive systems such as national critical information infrastructure, defense and military industry, and core government affairs, or to bypass national security defense systems. If obtained by non-state actors or malicious entities, they could endanger national security and the common security of humanity. The Chinese government implements necessary management over such technologies with the aim of building a solid security bottom line, preventing abuse and malicious use, ensuring AI remains under human control, and advancing international AI governance.
Second, to encourage open-sourcing, openness, cooperation, and sharing. The Chinese government is committed to making AI an important power source for promoting common prosperity and safeguarding common security. To support building an AI open-source innovation ecosystem, for technology exports such as open-sourcing of AI technologies, Chinese enterprises building and operating computing centers abroad, carrying out chip tape-out manufacturing abroad, conducting joint R&D using overseas resources, and participating in international standard setting, facilitation measures such as automatic licensing apply, basically achieving 'export upon filing'; for related technologies that have already entered the public domain through open-sourcing, publication of academic papers, and other means, it is proposed not to include them in the scope of export management, creating a favorable environment for international exchange and cooperation in AI technology.
It should be particularly pointed out that China advocates conducting AI dialogue and cooperation with all countries on the basis of mutual respect, equality, and mutual benefit, jointly addressing the risks of AI technology proliferation and loss of control. China supports Global South countries in strengthening AI capacity building, bridging the digital and intelligence divide, and promoting sustainable development. For technology exports needed for international cooperation on capacity building, technology sharing, and rule building among members of the World AI Cooperation Organization, more facilitation and support will be provided.
Q: After the Catalogue is formally implemented, do various forms of 'going overseas' such as localized deployment of AI large models abroad and providing model calls to foreign parties fall within the scope of technology export management?
A: At present, the 'going overseas' models of AI large models and application products are diverse; the core criterion for determining whether a specific 'going overseas' situation falls within the scope of technology export management is whether it involves the transfer of restricted technologies from within China to outside the country.
The scope of AI technologies included in regulation this time is precise and limited; after the relevant measures are formally implemented,the vast majority of AI technologies can still be exported freely. For AI technologies included in the restricted category, carrying out localized deployment abroad falls within the scope of technology export management; if deployment occurs in relatively controllable situations such as overseas self-owned clouds, facilitation measures such as automatic licensing may apply. For cases where enterprises use restricted large models to provide API calls externally, APP/web interactions, SaaS applications, or lightweight local on-device deployment, these do not fall within the scope of large model technology export management.