Shanghai Regulator Issues 16 Measures to Pilot Generative AI in Banking and Insurance
On September 24, the Shanghai Financial Regulatory Bureau issued 16 measures to promote AI in banking and insurance. The policy encourages vertical model development, data governance, and a pilot for customer-facing generative AI in controlled environments. It also supports a flexible regulatory framework with fault-tolerance mechanisms and requires risk management against AI hallucinations and deepfakes.
IllustrationEditorial responsibility
- No named human review is recorded for this page.
- Reports are grouped by semantic similarity and deterministic rules. Language models may assist titles, summaries, translation and cross-source analysis; the page reads the event directly, while its address stays stable when the title changes.
- Summary covers the current reports
Cross-source coverage
Reporting timeline
Shanghai regulator allows banks to gradually deploy customer-facing AI models in controlled environments
On September 24, the Shanghai Financial Regulatory Bureau issued a policy document titled 'Several Measures to Promote the Application of Artificial Intelligence in Shanghai's Banking and Insurance Industries,' containing 16 specific measures across three areas. The measures encourage financial institutions to deepen AI use in core business scenarios such as smart marketing, credit, claims processing, and risk control, and to cautiously explore financial AI agent construction. The bureau supports a 'rent-and-purchase' model for optimizing AI computing resources and building enterprise-level AI platforms. It requires institutions to establish governance structures, conduct model registration or filing for public-facing generative AI services, and implement high-risk application access controls to prevent technology abuse and digital formalism. Notably, the bureau will explore a pilot mechanism for generative AI large models in finance, seeking to include Shanghai in national pilot programs, allowing institutions to gradually deploy customer-facing large model applications in controlled environments. It also plans to study a tiered, inclusive regulatory framework with differentiated tolerance for varying risk levels, aiming to form a dynamic and collaborative AI governance system.
Read sourceShanghai Regulator Urges Banks, Insurers to Develop Vertical AI Models
On September 24, the Shanghai Financial Regulatory Bureau issued a set of measures to promote the integration of artificial intelligence with the banking and insurance industries. The document, titled 'Several Measures to Promote the Application of Artificial Intelligence in Shanghai's Banking and Insurance Industries,' encourages financial institutions to use AI to optimize marketing, customer service, and business processes such as intelligent credit and claims processing. It specifically supports institutions in independently developing vertical domain models, balancing priorities like real-time performance, accuracy, and explainability across different scenarios. The measures also recommend a procurement strategy combining general-purpose large models with industry-specific models, and the creation of a layered model architecture. On risk management, the policy requires institutions to build secure AI infrastructure using trusted chips, software, and data, establish data security mechanisms for AI training, and guard against risks such as deepfakes, AI hallucinations, and model security vulnerabilities. The goal is to evolve financial services from digital to intelligent capabilities while ensuring safety and controllability.
Read sourceShanghai Regulator to Pilot Generative AI Applications in Banking and Insurance
The Shanghai Financial Regulatory Bureau has issued a set of measures to promote the application of artificial intelligence in the banking and insurance sectors within the city. The document, titled 'Several Measures to Promote the Application of Artificial Intelligence in Shanghai's Banking and Insurance Industries,' outlines a plan to establish a pilot mechanism for the use of generative AI large models in finance. The bureau aims to have Shanghai included in a regional pilot program by the national Financial Regulatory Administration, which would allow financial institutions to gradually deploy large models directly facing customers in a controlled environment. This pilot would be coordinated with the cyberspace administration's registration and filing system for generative AI services. The regulator also proposes studying a flexible, tiered regulatory framework for AI that includes a fault-tolerance mechanism, with differentiated tolerance levels for events of varying risk, to foster a dynamic and collaborative AI governance structure.
Show 2 older updatesHide older updates
Shanghai Regulator Pushes Banks and Insurers to Strengthen Data Governance and AI Use
The Shanghai Financial Regulatory Bureau has issued a set of measures titled 'Several Measures to Promote the Application of Artificial Intelligence in Shanghai's Banking and Insurance Industries.' The document calls on financial institutions to enhance data governance by improving data management systems and quality control mechanisms. It encourages the integration of multi-source data to build high-quality datasets and establish full-process management standards for data collection, cleaning, standardization, and labeling. The bureau also urges institutions to build knowledge bases and knowledge maps, clarifying knowledge boundaries and relationships. A knowledge management system with a full lifecycle mechanism for knowledge generation, storage, updating, and retirement is recommended, along with content review and validity management. The measures further encourage the use of AI to extract knowledge and explore privacy computing and blockchain technologies for secure data sharing, subject to privacy protection and compliance requirements.
Read sourceShanghai Regulator Issues Measures to Promote AI Application in Banking and Insurance
On September 24, the Shanghai Financial Regulatory Bureau issued a set of measures to promote the application of artificial intelligence in the city's banking and insurance industries. The policy outlines several key initiatives. It encourages financial institutions to optimize the allocation of intelligent computing resources by adopting a 'rent-and-purchase' model to build hybrid cloud computing platforms, leasing national computing nodes or industry infrastructure for flexible and scalable computing power. The measures also support the development of vertical domain models, allowing institutions to balance priorities like real-time performance, accuracy, and interpretability across different scenarios. A strategy of using 'general large models as a foundation plus industry-specific models for deployment' is recommended, along with the creation of a one-stop large model application platform with scenario-based templates to lower the barrier to entry. Furthermore, the policy urges financial institutions to improve their AI evaluation systems, align with industry practices and risk control requirements, and participate in setting model assessment and certification standards.