Ep.3 "Who Regulates AI in China, and Can They Find Common Ground with the U.S.?"
China AI Regulation: A Cross-Departmental Matrix Governance
As AI regulations tighten worldwide, China has taken a different path from the EU. Instead of a centralized AI law, China relies on a complex web of agencies, each with its own oversight role. China’s AI regulation operates through inter-agency collaboration, where multiple government bodies coordinate policies, enforcement, and compliance oversight. Several key governmental bodies frequently collaborate on AI-related policies and enforcement:
Cyberspace Administration of China (CAC): The content regulator, responsible for overseeing deepfake technologies, generative AI, and algorithmic recommendation systems.
Ministry of Industry and Information Technology (MIIT): The industrial policy maker, supervising the commercialization of AI applications in sectors like autonomous driving and healthcare.
Ministry of Public Security (MPS): The security watchdog, focusing on facial recognition, AI-driven surveillance, and cybersecurity risks.
State Administration for Market Regulation (SAMR): The fair competition enforcer, preventing algorithmic discrimination and anti-competitive practices like price manipulation through big data.
Ministry of Science and Technology (MOST): The technology strategist, responsible for setting AI ethical standards and guiding fundamental research.
Two Regulatory Pillars: How China Governs AI
China’s AI regulatory framework operates on two primary axes:
Use-Case-Based Regulation → Focuses on AI applications across industries
Algorithmic Recommendation (2021) – Prevents information manipulation and big-data-driven price discrimination. (January 2022)
Deep Synthesis (2022) – Requires AI-generated face swaps and voice synthesis to be clearly labeled. (December 2022)
Generative AI (2023) – Introduces review mechanisms for ChatGPT-like models to ensure content compliance. (July 2023)
Industry-Specific Regulation → Focuses on AI risks within certain sectors
Autonomous Driving (July 2021)
Medical AI (July 2021)
For a deeper dive into China's AI regulatory landscape, you can read more here.
At first glance, China’s AI regulatory approach appears fragmented compared to the EU’s AI Act, which establishes a unified legal framework classifying AI into different risk categories. Instead of a sweeping, one-size-fits-all law, China has opted for a sector-specific, iterative regulatory model.
“But Why has Beijing chosen this gradual, decentralized approach over a comprehensive AI ACT?
One explanation lies in China’s broader regulatory philosophy—'small, fast, and flexible' (小、快、灵). This approach aligns with China’s policy tradition of first piloting regulations in key industries before expanding them nationwide. Unlike the EU, where a single AI law is designed to cover all sectors, China’s regulatory framework is shaped by its multi-agency governance model, where different ministries oversee AI applications based on their specific policy mandates. This decentralized structure allows China to rapidly test AI regulations while balancing industry growth and government oversight.
Pragmatic Regulation or Legal Uncertainty?
Beijing has adopted a case-by-case regulatory approach, ensuring that AI governance remains aligned with national security priorities and economic development goals, rather than being constrained by a single legal framework. However, this model is not without its shortcomings.
Flexibility vs. Stability: While sector-specific regulations allow for rapid policy adjustments, they may also create uncertainty for businesses, as compliance requirements can change unpredictably.
Government Control vs. Market Innovation: China’s multi-agency regulatory system ensures strict oversight, but the absence of a clear, unified AI law means that businesses must navigate a fragmented and potentially overlapping regulatory framework, which may hinder technological innovation. Additionally, due to the decentralized regulatory structure, certain AI applications in specific regions or industries may not receive adequate oversight, making it difficult to identify and mitigate potential risks in a timely manner.
International Alignment vs. Sovereign Regulation: The EU AI Act aims to establish a global standard for AI governance, whereas China’s decentralized approach keeps regulations highly localized, making cross-border compliance more complex.
Despite these concerns, discussions about enacting a unified AI law have not entirely disappeared. Since 2023, Chinese legal scholars and think tanks have debated the possibility of integrating AI regulations into a centralized framework. The **Model AI Law 1.0 (Expert Recommendation Draft)** proposes consolidating multiple AI-related regulations under a single national authority, drawing inspiration from the EU’s approach. However, while the State Council has included AI legislation in its 2023 and 2024 Legislative Work Plans, the National People's Congress (NPC)—China’s highest-ranking legislative body—has yet to establish a concrete timeline for advancing such a law, despite having the strongest legal authority within China’s legislative system.
Interesting Similarities Between AI Regulatory Models in China and the U.S.
Interestingly, when it comes to AI governance, China and the United States share several similarities in their regulatory models and legal frameworks.
Industry-Led Regulatory Approach
In the U.S., most AI governance is handled by sector-specific regulatory agencies, rather than through a single, overarching federal law. This approach mirrors China’s reliance on industry regulations and administrative guidelines for AI oversight. Examples include:
The Food and Drug Administration (FDA), which regulates AI-assisted medical products to ensure their safety and efficacy.
National Security as an Exception: Unlike other sectors, AI regulation in national security is tightly controlled by the U.S. government, which has broad authority over AI applications in military use. Additionally, the U.S. enforces export controls on advanced semiconductors, aiming to limit China’s ability to develop its own military AI capabilities.
The Role of the White House and Federal Government: Rather than enacting a unified AI law, the White House and federal agencies primarily serve as coordinators and policy drivers, influencing AI governance through multi-stakeholder discussions, policy goal-setting, and voluntary standards. This approach aligns with China’s strategy of gradual adjustments, policy trials, and industry-led governance.
Challenges in AI Legislation
U.S. Legislative Hurdles: According to Karman Lucero, a researcher at Yale Law School, the broad scope of the EU’s newly passed AI law would likely be too expansive to pass through both the House and Senate in the U.S.
Constitutional Constraints on AI Regulation: Recent rulings by the U.S. Supreme Court in Murthy v. Missouri (2024) and Moody v. NetChoice (2024) have reinforced the view that algorithms and AI-based content moderation fall under First Amendment protections as free speech. This interpretation sets a high threshold for regulatory intervention, making it significantly harder for the U.S. government to impose direct, comprehensive AI regulations.
Conclusion: Toward Practical International AI Cooperation
As AI technology continues to advance, both China and the U.S. will face mounting pressure to refine their regulatory approaches while addressing the global challenges posed by AI governance. While their strategies are not entirely aligned—China’s top-down, adaptive regulation versus the U.S.’s sector-driven, constitutionally constrained model—both countries recognize that AI’s risks and opportunities extend beyond national borders.
Given the interconnected nature of AI development, supply chains, and deployment, effective AI governance will require pragmatic international cooperation. While deep policy alignment between China and the U.S. remains unlikely, there is room for practical collaboration in areas such as AI safety research, risk management frameworks, and technical standards for AI transparency and accountability.
In the next episode of this series, we will explore the history of Track 1 and Track 1.5 dialogues between China and the U.S., examining past diplomatic and policy discussions on technology and governance. By looking at what worked, what didn’t, and the lessons learned, we can better understand how future AI-related exchanges can contribute to more effective international cooperation in this rapidly evolving field.