From 20 to 23 October 2025 the Fourth Plenary Session of the 20th Central Committee of the Communist Party of China was held in Beijing and adopted the Recommendations for the 15th Five-Year Plan for National Economic and Social Development. The Recommendations make a dual deployment on AI: development and governance.

The development side: digital-intelligent technology and “AI plus”

The Recommendations place acceleration of AI and other digital-intelligent technology innovation in a prominent position, calling for breakthroughs in basic theory and core technology, more efficient supply of compute, algorithms and data, and full implementation of the “AI plus” initiative.

This consolidates previously scattered policy threads into one main line: compute, algorithms and data are the factor base, “AI plus” is the application path, and core technology breakthroughs are the capability guarantee. See our notes on the State Council’s “AI plus” deployment and building an integrated national computing system.

Of particular note, the Recommendations list “embodied intelligence” as a growth point for future industries. This raises the combination of AI with robotics and the physical world to the level of industrial strategy, and the related product liability, safety standards and data rules will become new compliance topics. See our note on NVIDIA’s physical AI and embodied intelligence.

The governance side: four layers of rules

The governance wording deserves careful reading: strengthen AI governance, and improve relevant laws and regulations, policy systems, application norms and ethical guidelines.

The four terms correspond to four layers of rules: laws and regulations are hard constraints, policy systems are administrative rules, application norms are industry and scenario standards, and ethical guidelines are value orientation. Listing them together means future AI governance will not rest on a single statute but on multi-layer rule coordination.

The practical implication is that compliance work cannot watch only laws and regulations; it must track policy systems, application norms and ethical guidelines, which often precede legislation and become important references in enforcement and adjudication. See our notes on China’s AI regulatory system and the AI application ethics and security guideline.

Days before the Recommendations, the Standing Committee of the National People’s Congress adopted amendments to the Cybersecurity Law, writing an AI-specific clause into a foundational statute for the first time. See our note on the AI clause in the revised Cybersecurity Law.

The proximity of top-level design and legislative amendment shows the hierarchy of AI governance rising as a whole: from ministry rules to a structured system of law, policy, standards and ethics. Dedicated AI legislation is the expected next step. See our note on the AI legislation process.

Where companies can start

First, move the governance framework from responding to inspections towards internalised systems, reviewing existing coverage against the four rule layers.

Second, track where “AI plus” lands in your own industry, since application norms usually follow scenario policy.

Third, factor new forms such as embodied intelligence into forward compliance planning, especially product safety, liability allocation and data collection rules.

The significance of the Recommendations is not the duties added immediately, but the framework and cadence they set for AI governance over the next five years.

This article is based on the Recommendations for the 15th Five-Year Plan adopted at the Fourth Plenary Session of the 20th Central Committee and on public information. It is for general reference only and does not constitute legal advice.