.png){kind=logo}
October 2025
Shanghai AI-Industry Policy Landscape (2020-2025)
Following the State Council's "Opinions on Deepening the 'AI +' Action" (Guo Fa [2025] No. 11), Shanghai released in September 2025 the "Notice on Organising the 2025 Municipal 'AI +' Project Application," reaffirming its ambition to build an "AI Shanghai Highland." This memo, therefore, summarises Shanghai's AI-related policies issued during the past three years for investors and market players.
I. Key Policy Instruments
Since 2022, Shanghai has issued more than a dozen documents that can be grouped into two tiers: top-level system design and vertical-domain support measures. Together, they define the AI industry, lay down development principles, assign responsibilities, set support directions and call for the joint advancement of technology, industry, applications and governance.
A. Top-level Design
- Regulations on Promoting the Development of the AI Industry (Shanghai)
- Implementation Plan for "Model-Shaping Shanghai"
- Several Measures for Promoting Innovation of Large AI Models in Shanghai (2023-25)
- China (Shanghai) Pilot Free-Trade Zone Regulations (2025)
- Action Plan for Seizing the Digital-Economy Track and Boosting the Data-Element Industry (2023-25)
- Implementation Opinions of Shanghai People's Government on Further Leveraging the Capital Market to Foster High-Quality Development of Sci-Tech Enterprises
- Several Measures for Expanding AI Applications in Shanghai
B. Vertical-Domain Support
- Several Measures for Empowering Industrial Upgrading through Producer Services
- High-Quality Development Action Plan for Smart-Terminal Industry (2026-27)
- Action Plan for High-Quality and Innovative Development of Intelligent Robotics (2023-25)
- Several Measures for AI-Enabled Transformation of HR Service Industry
- Opinions on Further Promoting High-Quality Development of Venture Capital in Shanghai
- Action Plan for Empowering Industrial Upgrading via Producer Services (2024-27)
- Several Policy Measures for Healthy Development of the Online New Economy
- Work Plan for Medical AI Development (2025-27)
- Three-Year Action Plan for High-Quality Manufacturing Development (2023-25)
- Action Plan for Science & Technology Innovation in Elderly-Care (2024-27)
- Opinions on Boosting High-Quality Development of Producer Internet Platforms
II. Priority Support Areas
A. Foundational Technology & Platforms
- Large-model stack: L0 general models, L1 sector-specific base models, L2 scenario models; open-source, compression (distillation, pruning, quantisation).
- AI computing power: ultra-large autonomous clusters (GPU/ASIC/FPGA), city-level scheduling platform, training-inference integrated services.
- Data & corpora: high-quality domain corpora, multimodal acquisition/synthesis/annotation, data-sharing & revenue-sharing, public-data opening and data-element circulation.
B. Vertical Applications
- Manufacturing: smart factories, industrial large models, defect detection, process optimisation, denser industrial robots.
- Finance: smart risk-control, robo-advisory, anti-fraud models, large financial models, intelligent customer service.
- Healthcare: medical imaging, AI-assisted diagnosis, drug-discovery models, pre-diagnosis chatbots, rehab robotics.
- Education: AI teaching platforms, AI teaching assistants, personalised learning, education large models, digital humans.
- Urban governance: city-scale models, smart traffic, environmental monitoring, public-safety early-warning, city brain.
- Autonomous driving: end-to-end driving models, vehicle-road coordination, smart cockpits, automotive chips, robo-taxi/robo-truck.
- Embodied AI: humanoid robots, spatial intelligence, multimodal interaction, robot datasets, open-source simulation.
- AI for Science: molecular design, drug screening, materials R&D, scientific large models.
- Online new economy: smart search, content generation, digital anchors, intelligent customer service, platform intelligence upgrade.
III. Principal Support Instruments
A. Fiscal Incentives [1]
- Basic-theory & critical technology breakthroughs: up to RMB 50 million, ≤30% of verified investment.
- AI computing-power lease: up to 100% rent subsidy for ≤1 year (city + district).
- Model API or on-prem deployment: "model vouchers" up to RMB 5 million, ≤50% of contract value.
- Corpus procurement: "corpus vouchers" up to RMB 5 million, ≤30% of contract value.
- Benchmark demonstration scenarios: up to RMB 10 million, ≤20% of verified investment.
- First-version high-end software: up to RMB 20 million, ≤30% of contract value.
- Industrial innovation service platforms: up to RMB 20 million, ≤50% of verified investment.
- Large-scale embodied-robot deployment: 5% of contract value, ceiling RMB 5 million.
B. Infrastructure & Platforms
- Manufacturing Innovation Centre (humanoid robots), testing & pilot-validation centre, General-Purpose Robotics Research Institute. [2]
- Scale production bases for smart terminals: higher subsidies for tech retrofits, "smart factory" incentives, producer-service support (R&D, logistics, testing, supply-chain), pilot platforms, software-hardware adaptation centres. [3]
C. Talent & Finance
- Talent: household-registration (hukou), housing, schooling, medical care; interdisciplinary AI majors, industry-training bases. [4]
- Finance: industrial transformation funds, pilot funds, state-owned capital parent funds guiding private capital; M&A support for key links. [5]
D. Data & Standards [6]
- Data: dynamic list of public-data releases in healthcare, transport and government; support for data trading, asset registration, and IP pilots.
- Standards: encourage leadership in national/industry/local AI standards; algorithm-model filing, evaluation, and trading; "Shanghai Standard" AI certification.
IV. Conclusion
Shanghai has built a full-chain policy framework—basic capacity → vertical application → industrial ecosystem → factor supply—anchored in the national "AI +" strategy, underpinned by the Municipal AI Industry Regulations and centred on the "Model-Shaping Shanghai" project. The orientation is clear, instruments are diversified, fiscal fire-power is large and application scenarios are open, offering an attractive environment for AI companies, research institutes and developers.
[1] See "Several Measures for Further Expanding the Application of Artificial Intelligence in Shanghai."
[2] See "Action Plan for High-Quality and Innovative Development of the Intelligent Robotics Industry in Shanghai (2023-2025)."
[3] See "Action Plan for High Quality Development of Intelligent Terminal Industry in Shanghai (2026-2027)."
[4] See "Regulations on Promoting the Development of Artificial Intelligence Industry in Shanghai and Action Plan for Promoting High Quality Innovation and Development of Intelligent Robot Industry in Shanghai (2023-2025)."
[5] See "Action Plan for High Quality Development of Intelligent Terminal Industry in Shanghai (2026-2027)."
[6] See "Regulations on Promoting the Development of Artificial Intelligence Industry in Shanghai."
I. Key Policy Instruments
Since 2022, Shanghai has issued more than a dozen documents that can be grouped into two tiers: top-level system design and vertical-domain support measures. Together, they define the AI industry, lay down development principles, assign responsibilities, set support directions and call for the joint advancement of technology, industry, applications and governance.
A. Top-level Design
- Regulations on Promoting the Development of the AI Industry (Shanghai)
- Implementation Plan for "Model-Shaping Shanghai"
- Several Measures for Promoting Innovation of Large AI Models in Shanghai (2023-25)
- China (Shanghai) Pilot Free-Trade Zone Regulations (2025)
- Action Plan for Seizing the Digital-Economy Track and Boosting the Data-Element Industry (2023-25)
- Implementation Opinions of Shanghai People's Government on Further Leveraging the Capital Market to Foster High-Quality Development of Sci-Tech Enterprises
- Several Measures for Expanding AI Applications in Shanghai
B. Vertical-Domain Support
- Several Measures for Empowering Industrial Upgrading through Producer Services
- High-Quality Development Action Plan for Smart-Terminal Industry (2026-27)
- Action Plan for High-Quality and Innovative Development of Intelligent Robotics (2023-25)
- Several Measures for AI-Enabled Transformation of HR Service Industry
- Opinions on Further Promoting High-Quality Development of Venture Capital in Shanghai
- Action Plan for Empowering Industrial Upgrading via Producer Services (2024-27)
- Several Policy Measures for Healthy Development of the Online New Economy
- Work Plan for Medical AI Development (2025-27)
- Three-Year Action Plan for High-Quality Manufacturing Development (2023-25)
- Action Plan for Science & Technology Innovation in Elderly-Care (2024-27)
- Opinions on Boosting High-Quality Development of Producer Internet Platforms
II. Priority Support Areas
A. Foundational Technology & Platforms
- Large-model stack: L0 general models, L1 sector-specific base models, L2 scenario models; open-source, compression (distillation, pruning, quantisation).
- AI computing power: ultra-large autonomous clusters (GPU/ASIC/FPGA), city-level scheduling platform, training-inference integrated services.
- Data & corpora: high-quality domain corpora, multimodal acquisition/synthesis/annotation, data-sharing & revenue-sharing, public-data opening and data-element circulation.
B. Vertical Applications
- Manufacturing: smart factories, industrial large models, defect detection, process optimisation, denser industrial robots.
- Finance: smart risk-control, robo-advisory, anti-fraud models, large financial models, intelligent customer service.
- Healthcare: medical imaging, AI-assisted diagnosis, drug-discovery models, pre-diagnosis chatbots, rehab robotics.
- Education: AI teaching platforms, AI teaching assistants, personalised learning, education large models, digital humans.
- Urban governance: city-scale models, smart traffic, environmental monitoring, public-safety early-warning, city brain.
- Autonomous driving: end-to-end driving models, vehicle-road coordination, smart cockpits, automotive chips, robo-taxi/robo-truck.
- Embodied AI: humanoid robots, spatial intelligence, multimodal interaction, robot datasets, open-source simulation.
- AI for Science: molecular design, drug screening, materials R&D, scientific large models.
- Online new economy: smart search, content generation, digital anchors, intelligent customer service, platform intelligence upgrade.
III. Principal Support Instruments
A. Fiscal Incentives [1]
- Basic-theory & critical technology breakthroughs: up to RMB 50 million, ≤30% of verified investment.
- AI computing-power lease: up to 100% rent subsidy for ≤1 year (city + district).
- Model API or on-prem deployment: "model vouchers" up to RMB 5 million, ≤50% of contract value.
- Corpus procurement: "corpus vouchers" up to RMB 5 million, ≤30% of contract value.
- Benchmark demonstration scenarios: up to RMB 10 million, ≤20% of verified investment.
- First-version high-end software: up to RMB 20 million, ≤30% of contract value.
- Industrial innovation service platforms: up to RMB 20 million, ≤50% of verified investment.
- Large-scale embodied-robot deployment: 5% of contract value, ceiling RMB 5 million.
B. Infrastructure & Platforms
- Manufacturing Innovation Centre (humanoid robots), testing & pilot-validation centre, General-Purpose Robotics Research Institute. [2]
- Scale production bases for smart terminals: higher subsidies for tech retrofits, "smart factory" incentives, producer-service support (R&D, logistics, testing, supply-chain), pilot platforms, software-hardware adaptation centres. [3]
C. Talent & Finance
- Talent: household-registration (hukou), housing, schooling, medical care; interdisciplinary AI majors, industry-training bases. [4]
- Finance: industrial transformation funds, pilot funds, state-owned capital parent funds guiding private capital; M&A support for key links. [5]
D. Data & Standards [6]
- Data: dynamic list of public-data releases in healthcare, transport and government; support for data trading, asset registration, and IP pilots.
- Standards: encourage leadership in national/industry/local AI standards; algorithm-model filing, evaluation, and trading; "Shanghai Standard" AI certification.
IV. Conclusion
Shanghai has built a full-chain policy framework—basic capacity → vertical application → industrial ecosystem → factor supply—anchored in the national "AI +" strategy, underpinned by the Municipal AI Industry Regulations and centred on the "Model-Shaping Shanghai" project. The orientation is clear, instruments are diversified, fiscal fire-power is large and application scenarios are open, offering an attractive environment for AI companies, research institutes and developers.
[1] See "Several Measures for Further Expanding the Application of Artificial Intelligence in Shanghai."
[2] See "Action Plan for High-Quality and Innovative Development of the Intelligent Robotics Industry in Shanghai (2023-2025)."
[3] See "Action Plan for High Quality Development of Intelligent Terminal Industry in Shanghai (2026-2027)."
[4] See "Regulations on Promoting the Development of Artificial Intelligence Industry in Shanghai and Action Plan for Promoting High Quality Innovation and Development of Intelligent Robot Industry in Shanghai (2023-2025)."
[5] See "Action Plan for High Quality Development of Intelligent Terminal Industry in Shanghai (2026-2027)."
[6] See "Regulations on Promoting the Development of Artificial Intelligence Industry in Shanghai."
本網頁上所有上海理慈法律新知資料內容（「內容」）均屬上海理慈律師事務所所有。上海理慈保留所有權利，除非獲得上海理慈事前許可外，均不得以任何形式或以任何方式重製、下載、散布、發行或移轉本網頁上之內容。
所有內容僅供作參考且非為特定議題或具體個案之法律或專業建議。所有內容未必為最新法律及法規之發展，上海理慈及其編輯群不保證內容之正確性，並明示聲明不須對任何人就信賴使用本網頁上全部或部分之內容，而據此所為或經許可而為或略而未為之結果負擔任何及全部之責任。撰稿作者之觀點不代表上海理慈之立場。如有任何建議或疑義，請與上海理慈聯繫。