The rapid advancements in AI and digital therapeutics have brought transformative changes to the health care landscape in China. From AI-powered diagnostics to software-as-a-medical-device (SaMD), these innovations promise improved patient outcomes and operational efficiencies. They also make it imperative for stakeholders in the life sciences sector to stay informed about the regulatory landscape in China.

The regulatory starting line: "Intended purpose" as gatekeeper

The primary factor determining whether a digital health product falls under China's medical device regulations is its "intended purpose." This distinction separates regulated medical devices from general wellness applications. For example, an algorithm that analyzes medical device data to provide diagnostic decision support would be subject to the medical device regulations.

The National Medical Products Administration (NMPA) adopts a risk-based classification system to categorize and regulate medical devices. Most AI-powered SaMD and digital therapeutics are classified as Class II or III medical devices depending on their core functions and the maturity of the underlying algorithms. Key points from NMPA guidelines, such as the Guiding Principles for the Classification and Definition of AI-Based Medical Software Products, indicate that software with low algorithmic maturity supporting clinical decision-making, like treatment planning, is typically classified as Class III. Conversely, software used solely for non-decision-making purposes, such as data processing for clinical reference, is generally classified as Class II. This risk-based approach ensures a proportionate level of regulatory scrutiny.

Enhanced data protection for health care AI

Beyond NMPA regulations, health care AI products must comply with China's comprehensive data protection framework, including the Personal Information Protection Law, Data Security Law, and Cybersecurity Law. Given that health care AI depends on sensitive patient data, any collection and use of such data must be legitimate, necessary, and aligned with the principle of data minimization. When training AI models with patient data in China, adherence to these requirements is essential.

Future outlook: A maturing regulatory environment

China's regulatory landscape for digital health is evolving rapidly. The government is actively promoting innovation through initiatives such as the 2025 Implementation Opinions on Promoting and Regulating the Application Development of "Artificial Intelligence + Healthcare," while simultaneously strengthening the legal framework to address emerging challenges, such as algorithm transparency and ethical standards.

Success in the Chinese market necessitates a compliance-by-design approach. This entails early engagement in assessment to clarify product classification, the implementation of a robust quality management system, and the integration of data protection principles throughout the product lifecycle. These considerations are not merely regulatory hurdles; they are strategic imperatives essential for sustainable growth.