In June 2025 the Ministry of Industry and Information Technology studied how to advance the AI industry and enable new industrialisation, and at the subsequent meeting of the leading group on industrial-digital integration required an AI plus manufacturing initiative: faster intelligent upgrading of key industries, and an “upgraded” version of smart manufacturing.
General and industry-specific models side by side
The meeting called for a coordinated layout of general-purpose and industry-specific large models. That phrasing shows policy is not pursuing a single technology path: general models supply foundation capability, industry models solve concrete problems, and the two have to work together.
For industrial companies, the question is therefore not which model to use, but which type of model to use at which step.
How application advantages are built
The meeting required large models to be deployed in key manufacturing industries, with “application advantage” as the objective. This aligns with the overall AI plus initiative, which uses applications to pull innovation forward. See our note on the State Council opinion on the AI plus initiative.
Security and ethics named together
Notably, the deployment discussion also referred to balancing development and security, requiring progress on deep synthesis detection technology and faster drafting of measures for AI technology ethics management.
Placing detection technology alongside ethics rules reflects a governance approach: use technical means to identify synthetic content, and use rules to make clear what may and may not be done. See our notes on the basic security requirements for generative AI services and on the AI application ethics and security guideline.
Where companies can start
First, break AI plus manufacturing down to specific process steps, identifying where model assistance is appropriate and where a human must remain responsible.
Second, watch the lawfulness of data sources for industry models. Industrial data often engages trade secrets and personal information. See our note on the action plan for building high-quality industry datasets.
Third, build ethics review into the product process, particularly where automated decision-making and workforce management are involved. See our notes on the AI plus manufacturing action plan and on the 2025 AI industry and industrialisation conference.
The value of manufacturing scenarios is that they push AI from demonstration towards delivery. Whether delivery happens depends on whether compliance capability can keep up with the pace of the production line.
This article is based on MIIT meetings on advancing the AI industry and enabling new industrialisation and public information and is for reference only; it does not constitute legal advice.