Artificial Intelligence and Freedom of Expression in China

This chapter examines China’s regulatory approach to generative AI and its implications for freedom of expression. While China has not enacted a stand-alone law on AI, it has developed a comprehensive patchwork of administrative measures and platform obligations that form a de facto regime of control over AI-generated content. Key frameworks include cybersecurity laws, administrative rules governing generative AI services, and interrelated rules targeting deep synthesis technologies and algorithmic recommendation systems. These frameworks collectively impose ideological, political, and technical constraints on generative AI systems, requiring alignment with socialist core values, national security objectives, and censorship norms. Platforms and developers are obliged to prescreen, filter, and monitor AI outputs, including for the purpose of combating copyright infringement, defamation, obscenity, misinformation, and any content deemed politically sensitive. Importantly, vague legal concepts such as “public order” and “social morality” are interpreted expansively to enable broad discretionary enforcement. China treats generative AI as a uniquely high-risk technology and imposes heightened liability and oversight regimes that significantly constrain its expressive potential. Although no formal distinction is made between human- and AI-generated content in law, AI outputs are often subjected to greater scrutiny due to their perceived uncontrollability and scale. Regulations also target specific use cases — such as AI-generated political commentary, sexually suggestive content, or satirical speech — under the guise of maintaining ethical standards or information security. Overall, China’s regulatory model emphasizes anticipatory censorship and political control, offering little room for protective measures like liability exemptions, independent oversight, or rights-based challenges. Rather than empowering users or protecting freedom of expression, the governance of generative AI in China reinforces existing authoritarian speech controls and extends them into emerging technological domains.