The Ministry of Industry and Information Technology (MIIT) recently launched an initiative on building industrial data foundation to pilot the construction of datasets of high-quality industries empowered by AI.

With the main focus on promoting the development and utilization of industrial data, key industries of the manufacturing sector will be selected, and category-specific policies will be introduced based on leading enterprises of key industries, platform institutions, advanced manufacturing clusters and city-level pilot programs for the digital transformation of small and medium-sized enterprises.

By the end of 2026, China will establish a number of industrial data consortia, build trusted interconnection platforms for key industries, gather a wealth of industrial data resources, tackle a batch of key data technology challenges, develop a set of industrial data standards, and create several high-quality, standardized and tradable industrial datasets.

Support will be provided for the application of industrial large models, industrial intelligent agents and other applications. Effective approaches, innovative mechanisms and experience models for the efficient collection and processing, trusted circulation and aggregation, and in-depth integrated application of industrial data will be summarized and formed to empower industries to improve quality, reduce costs and increase efficiency.

Key tasks of the action include establishing a trusted interconnection platform for industrial data, an industrial data resource repository, a repository for key data technology breakthroughs, an industrial data standards library, a repository of high-quality industry datasets and a portfolio of data application scenarios.

Efforts will be made to integrate data infrastructure capabilities, including the Industrial Internet, big data centers, computing power centers, data governance platforms, data engineering capability platforms and industry-specific trusted data spaces to build a trusted data interconnection platform for key industries. Industrial data production and training bases will also be established to provide data collection in simulation environments, specialized data annotation and data synthesis for extreme scenarios.

Focusing on key stages such as R&D and design, pilot testing and validation, production and manufacturing, operational management and industrial collaboration, the industry-specific large model applications and industrial AI agent development will be promoted to empower the industry in improving quality, reducing costs and enhancing efficiency.

MIIT also called for financial support through local incentives, the development of an industrial data open-source section based on national AI open-source communities, and closer collaboration between universities and enterprises.