At a press conference themed “Launch of the 15th Five‑Year Plan” held by the State Council Information Office on July 29, 2026, Rui Wenbiao, Deputy Commissioner of the China National Intellectual Property Administration (CNIPA), stated that CNIPA will “guide the construction of patent pools in the artificial‑intelligence sector, facilitate the transfer of more high‑quality patents from laboratories to industrial chains and empower industrial development”. This statement has drawn wide‑ranging industry attention. What exactly is a patent pool? What is the significance of building one for AI? What does it mean for small‑and‑medium‑sized enterprises (SMEs)?

1. What Is a Patent Pool?

Simply put, a patent pool is a model under which multiple patent‑holders jointly operate their patents. According to the Guidelines for the Construction and Operation of Patent Pools jointly issued by CNIPA and five other governmental departments, a patent pool refers to “a patent‑utilization model in which two or more patent‑holders entrust one party or a third‑party operation‑management institution via agreement to jointly operate patents within a certain technical field, conducting cross‑licensing, one‑stop licensing and related services”.

To draw an analogy: traditional patent licensing resembles negotiating deals one‑by‑one, which is time‑consuming and labor‑intensive. A patent pool acts like a “patent supermarket”. Patent owners contribute patents into the pool, and the pool provides unified one‑stop licensing services to the outside world. Users only need to deal with the pool to obtain authorization for required patents. Major functions of patent pools include integrating patent resources, lowering transaction costs for licensing and promoting industrial application of patented technologies. Its construction and operation follow four principles: market‑orientation, interest balance, openness and non‑discrimination.

Internationally, there are well‑established precedents such as the 6C patent pool for DVD technology and MPEG‑LA in communications. Nevertheless, in the emerging AI field, patent‑pool practice is still at the exploratory stage worldwide.

2. Why Does the AI Sector Need Patent Pools?

In April 2025, Shen Changyu, Commissioner of CNIPA, noted that China has become “the world’s largest holder of artificial‑intelligence patents, accounting for 60 % of global AI patents”. Despite the large patent volume, many patents remain “locked‑up in laboratories”. Universities and research institutions hold patents but lack transformation channels, while enterprises have technical demands yet cannot find available technologies.

AI patent pools serve as an institutional tool to resolve such supply‑demand mismatch. Specifically, they deliver three core values.

First, breaking the “patent jungle”. AI technology spans algorithms, computing power, data and application scenarios, with massive overlapping patents. Individual enterprises can hardly sort out all ownership relationships. Centralized licensing via patent pools helps enterprises bypass “patent minefields”.

Second, cutting transaction costs. For AI enterprises, negotiating patent licenses separately is both time‑consuming and costly. The one‑stop‑licensing model of patent pools substantially reduces transaction expenses.

Third, accelerating achievement transformation. CNIPA clarifies that the goal of patent‑pool construction is “to bring more high‑quality patents from laboratories into industrial chains”. In Qingyuan High‑tech Zone, the first local AI patent pool was launched in April 2025, incorporating 33 high‑value patents covering big data, large‑model and encryption technologies. Mechanisms such as licensing and cross‑sharing lower R&D costs for SMEs and attract high‑end AI projects.

3. Opportunities and Potential Challenges Faced by SMEs

Opportunities

Lower market‑entry threshold. In the past, SMEs often faced insurmountable patent barriers set by large‑scale market players. Through centralized licensing and reasonable pricing, patent pools make key technologies accessible to SMEs at affordable costs. As Wang Jiewan, Deputy Director of Qingyuan High‑tech Zone Operation Center, commented: “Through licensing authorization and cross‑sharing mechanisms, the patent pool reduces R&D costs for SMEs and attracts high‑end AI projects, generating innovation agglomeration effects.”

Mitigation of infringement risks. Patent litigations in the AI sector are growing in frequency. Once sued, SMEs usually cannot afford heavy litigation costs and damage awards. Obtaining licenses from or joining a patent pool functions like a “patent insurance”.

Challenges

Reasonableness of licensing fees. Licensing rates are set by initiators or operating entities. Excessively high pricing may bar SMEs from access. Therefore, the Guidelines for the Construction and Operation of Patent Pools require full consultation with potential licensees when determining or adjusting royalty rates to balance interests of all stakeholders.

Risk of new monopolies. Although patent pools reduce transaction costs, they may leverage joint advantages to eliminate or restrict competition. The Guidelines mandate third‑party assessment on validity and necessity of pooled patents, alongside anti‑monopoly compliance self‑inspections.

To join or to wait‑and‑see? SMEs shall assess their own technical strength and patent value. They may participate as patent‑holders to share revenue, or act as licensees to obtain authorized technologies. Different positions call for distinct strategies.

4. The Future Landscape of AI Patent Pools

The construction of AI patent pools has long been on the policy agenda. During the 2025 Two Sessions, Shen Changyu proposed to “promote the establishment of China’s patent pool for large AI models”. Afterwards, Hangzhou High‑tech Zone (Binjiang) started building its industrial AI patent pool, and Qingyuan High‑tech Zone launched its local AI patent pool. AI patent pools are evolving from conceptual design to real‑world practice, progressing from local pilots toward national guidance.

During the 15th Five‑Year‑Plan period, CNIPA will also “deepen international cooperation on AI‑related intellectual property and actively participate in the formulation of relevant international rules, technical standards and governance systems”. This means AI patent pools are not merely domestic industrial instruments; they may become an important lever for China to engage in global AI IP governance.

An AI patent pool is a double‑edged sword. Well‑designed, it can bridge the “last‑kilometer” gap between innovation chains and industrial chains. Poorly managed, it risks becoming a new tool for monopolistic conduct. Success hinges on sound institutional design, transparent operation and effective supervision. For AI‑related enterprises, passive waiting is not advisable. It pays to understand relevant rules in advance. After all, in patent‑driven competition, those who understand the rules first are more likely to win.