In December 2025 the Shanghai Municipal People’s Government General Office issued the Implementation Plan for Strengthening the Open Source System in Shanghai (Hu Fu Ban [2025] No. 33), elevating open source system building to an independent item of local industrial policy.

Targets and structure

The plan calls for accelerating the building of an autonomous open source system, releasing open source development potential and turning Shanghai into a nationally serving, internationally competitive open source innovation source and development highland.

The stage targets for 2027 are: an initial open source system with complete technical support, a sound service system, a reasonable talent structure and a prosperous open source industry; one to two internationally influential open source communities; 100 open-source commercialisation companies; more than 200 quality open source projects; and a concentration of open source talent.

From the published wording, the plan builds a “technology, talent, enterprise” trinity open source innovation ecosystem — corresponding to the supply of open source projects, the participants in open source activity, and the entities that realise open source value.

Why open source became a local policy lever

Open source has a special place in the AI industry. Large-model training is costly and iterates fast, so open source becomes an important route for technology diffusion and ecosystem building: companies open source models to attract developers and build standards influence, then monetise through services.

But cultivating an open source ecosystem suffers from classic market failure — a single company struggles to bear the long-term investment in community operations, infrastructure and talent training alone. The logic of local policy intervention is to fill that externality with public resources.

Nationally, Shanghai is not alone in this. In April 2026 Guangdong’s judicial and industry communities jointly released an AI open source ecosystem consensus, responding to the legal questions of open source ecosystems from a judicial and governance angle. See our note on the AI open source ecosystem consensus.

Three implications for AI open source companies

First, the open-source commercialisation path gains policy backing. The plan explicitly targets “100 open-source commercialisation companies”, placing the “open source plus commercial services” path within policy support. For startups built on open source models and monetising through enterprise services, this is a clear positive signal. See our notes on Kimi K2.5 and Kimi K2.6.

Second, open source is not immunity from liability. Open source licence commercial terms, downstream users’ compliance duties and third-party rights in open source code are all legal issues open source companies must handle. Policy support for open source does not remove responsibility on licences and IP.

Third, community governance rules need institutionalising. When a company becomes the leading party of an open source community, its governance rules — contributor agreements, trademark policy, licence change procedures — acquire a quasi-public character. Companies should fold community governance into their compliance systems.

Connection to the AI governance mainline

Open source and AI governance are in tension: open source lowers the technical threshold and also lowers regulatory traceability. The Shanghai plan leans to the “development” side, while governance requirements — including filing and labelling duties for downstream users of open source models — are set by national rules.

Companies need to understand both lines: policy encourages a thriving open source ecosystem, but deployers of open source models still have to discharge algorithm filing, generative AI service filing and content labelling duties. See our algorithm filing service.

The value of the Shanghai plan is elevating “open source” from a spontaneous company activity to industrial infrastructure requiring systematic construction. For AI companies, that is both a resource opportunity and a test of their open source governance capability.


This article is based on the Implementation Plan for Strengthening the Open Source System in Shanghai (Hu Fu Ban [2025] No. 33) and public information. It is for general reference only and does not constitute legal advice.