China has mandated that all companies engaged in artificial intelligence activities establish internal "AI ethics review committees" under new rules released by Beijing on Thursday, effective immediately. The directive, jointly issued by 10 government bodies including the Ministry of Industry and Information Technology, the National Development and Reform Commission, and the Chinese Academy of Sciences, applies not only to major tech firms like Alibaba and Baidu but also to universities, research institutions, and health organizations .

What Will These Ethics Committees Actually Review?

The new committees are tasked with evaluating AI-related activities across several critical dimensions. They must assess the impact of AI systems on human wellbeing, examine the fairness of AI algorithms, and determine whether AI systems are "controllable" and "explainable" . This represents a significant expansion of oversight mechanisms that Beijing first introduced in 2023, when the government established a unified science and technology ethics review system requiring companies engaged in "high-risk" activities to conduct ethical reviews before launching projects.

The 2023 system specifically targeted organizations developing AI models that "influence social discourse" or involve "highly autonomous decision-making." However, that earlier framework faced criticism for its ambiguity regarding operating scope and review standards, as well as the lack of effective enforcement mechanisms . The new mandate appears designed to address these shortcomings by making ethics reviews mandatory across the entire AI sector rather than only for high-risk applications.

How to Prepare Your Organization for AI Ethics Review Requirements

• Establish a Dedicated Committee: Form an internal ethics review committee with representatives from technical, legal, and business teams who can evaluate AI projects before deployment and ensure compliance with the new Beijing requirements.
• Document Algorithm Fairness Assessments: Create standardized processes for testing AI algorithms for bias and fairness, including regular audits of model outputs across different demographic groups and use cases.
• Develop Explainability Frameworks: Implement systems that allow your organization to document and demonstrate how AI systems make decisions, ensuring they meet the "explainable" standard required by the new rules.
• Build Controllability Safeguards: Design technical and operational controls that allow human oversight of AI systems, including the ability to intervene, pause, or override automated decisions when necessary.

The timing of this mandate is significant. Since 2022, major Chinese tech firms including Alibaba and Baidu have already established internal science and technology ethics review committees, suggesting that Beijing's new rules formalize practices that leading companies have voluntarily adopted . However, making these committees mandatory across all organizations signals that the government believes voluntary compliance is insufficient to manage AI risks at scale.

The emphasis on "controllability" and "explainability" reflects Beijing's broader policy approach to AI governance. Rather than restricting AI development, Chinese policymakers are attempting to ensure that rapid AI progress continues in what they describe as a "healthy" manner amid growing consumer and enterprise adoption . This contrasts with some Western regulatory approaches that focus more heavily on restricting certain AI applications or imposing strict liability frameworks.

Why Is China Tightening AI Oversight Now?

The new mandate arrives at a moment when AI adoption in China is accelerating across sectors. Large language models (LLMs), which are AI systems trained on vast amounts of text data to generate human-like responses, have proliferated in Chinese applications ranging from customer service to content creation. As these systems become more integrated into critical decision-making processes, the government appears concerned that without structured oversight, AI systems could cause unintended harms.

The requirement that committees assess whether AI systems "influence social discourse" suggests Beijing is particularly focused on content-related AI applications, including recommendation algorithms and generative AI tools that could shape public opinion or spread misinformation. By requiring internal review before deployment, the government aims to catch potential problems before they reach users at scale.

The inclusion of universities, research bodies, and health institutions in the mandate indicates that Beijing views AI ethics as a concern across all sectors, not just commercial technology companies. Healthcare AI systems that make diagnostic or treatment recommendations, for example, would now require internal ethics review to ensure fairness and explainability before clinical deployment.

For global technology companies operating in China or considering expansion into the Chinese market, the new rules represent an important regulatory development. Organizations will need to invest in ethics review infrastructure and ensure that their AI systems meet Beijing's standards for controllability and explainability. The mandate also signals that China intends to maintain a distinct regulatory approach to AI governance, one that emphasizes internal corporate responsibility and government oversight rather than relying solely on market forces or individual liability .