Ali Ikhwan
The Supreme People’s Court (SPC) of the People’s Republic of China has officially released a comprehensive set of judicial guidelines aimed at regulating the rapidly evolving landscape of artificial intelligence (AI). Titled "Opinions on Lawfully Handling Civil Cases Involving Artificial Intelligence," the document serves as a foundational framework for courts at all levels to navigate the complex legal challenges posed by generative AI, autonomous systems, and algorithmic decision-making. By integrating the principles of the Civil Code, the Cybersecurity Law, the Data Security Law, and the Personal Information Protection Law, the SPC aims to provide a predictable legal environment that fosters "new quality productive forces" while safeguarding individual rights and national security.
Background and Context: The Rise of AI Litigation
The issuance of these guidelines comes at a critical juncture in China’s technological trajectory. As the nation transitions into a digital-first economy, AI has become a primary driver of industrial transformation. However, the proliferation of AI technologies has outpaced traditional legal statutes, leading to a surge in novel disputes. In recent years, Chinese internet courts in Beijing, Hangzhou, and Guangzhou have seen a marked increase in cases involving AI-generated content (AIGC) copyright, deepfake-related personality infringements, and liability disputes concerning autonomous driving.
The SPC’s new Opinions represent a strategic effort to move beyond ad-hoc rulings toward a standardized judicial approach. The guidelines emphasize a "human-centric" philosophy, ensuring that technological advancement does not come at the expense of human dignity or social ethics. They specifically address the dual necessity of promoting innovation and maintaining high-level security, reflecting the government’s broader goal of making China a global leader in AI governance by 2030.
Chronology of AI Regulatory Development in China
The 2026 Opinions are the culmination of nearly a decade of regulatory evolution:
- July 2017: The State Council issued the "New Generation Artificial Intelligence Development Plan," establishing a three-step roadmap for AI leadership.
- January 2021: The Civil Code of the People’s Republic of China took effect, providing the basic legal vocabulary for personality rights and tort liability relevant to digital environments.
- November 2021: The Personal Information Protection Law (PIPL) was enacted, creating strict requirements for automated decision-making and data processing.
- August 2023: The "Interim Measures for the Management of Generative Artificial Intelligence Services" were implemented, marking one of the world’s first dedicated regulations for LLMs (Large Language Models).
- September 2026: The Supreme People’s Court issues the current Opinions to provide specific trial guidance for the judiciary, bridging the gap between high-level legislation and courtroom practice.
Protection of Personality Rights: Voice Cloning and Digital Resurrection
One of the most significant sections of the Opinions deals with the protection of personality rights in the age of generative AI. The SPC has addressed the controversial practice of "AI voice cloning" and the "digital resurrection" of deceased individuals.
Under the new guidelines, using AI to process a natural person’s name, image, or voice without explicit consent is strictly prohibited, unless otherwise provided by law. The court clarified that the "digital image" and "synthetic voice" of a person are extensions of their personality rights. In cases of "digital resurrection"—where AI is used to recreate the likeness or personality of a deceased person—the SPC ruled that the close relatives of the deceased have the right to seek legal redress if the AI representation infringes upon the deceased’s name, image, or reputation.
Furthermore, the guidelines tackle "human flesh searching" and "online doxing" facilitated by AI. Courts are instructed to rule against entities that use AI to track, analyze, or leak private information such as phone numbers, social media accounts, or physical locations, labeling such actions as direct infringements on the right to privacy.
Product Liability and Autonomous Systems
The Opinions provide much-needed clarity on the liability of autonomous vehicles and AI-driven products. As self-driving cars move from testing phases to public roads, the question of "who is at fault" during an accident has become a central legal dilemma.
The SPC established a tiered liability framework:
- Product Defects: If an accident is caused by a defect in the AI system or hardware, the manufacturer or seller is held liable under the Product Quality Law and the Civil Code.
- Driver Intervention: In vehicles equipped with "assisted driving" functions, the court will analyze the degree of human intervention. If a driver fails to take control when prompted by the system, liability may be shared between the human operator and the technology provider.
- Data Transparency: To ensure fair trials, the SPC has empowered courts to demand that manufacturers and operators provide "truthful and complete" data from event data recorders (EDRs) and autonomous driving logs. This "black box" transparency is essential for reconstructing the facts of an accident.
Intellectual Property and AI-Generated Content (AIGC)
The intellectual property (IP) section of the Opinions addresses the "originality" of AI-generated works. The SPC maintains that for a work to be protected under the Copyright Law, it must reflect a degree of human intellectual contribution.
If a natural person uses AI as a tool—providing specific prompts, structures, and creative direction—the resulting output may be recognized as a protectable work. However, purely machine-generated content without human creative input does not qualify for copyright.
The guidelines also address the "fair use" of data in training AI models. While the SPC encourages the development of open-source software and the lawful use of public data, it warns that the unauthorized use of copyrighted materials for model training must not "unreasonably harm the legitimate interests of the rights holders." This signals a move toward more stringent requirements for AI developers to disclose their training data sources.
Supporting Data and Market Impact
The necessity for these guidelines is underscored by the scale of China’s AI market. According to data from the China Academy of Information and Communications Technology (CAICT), the core AI industry in China was valued at over 600 billion RMB in 2024, with a projected annual growth rate of 18%.
By early 2026, the number of AI-related civil disputes in China increased by an estimated 45% year-on-year. Intellectual property disputes accounted for 30% of these cases, followed by personal information protection (25%) and consumer rights (20%). The SPC’s intervention is expected to stabilize the market by reducing "litigation risk premiums" for tech companies while providing a clear path for consumer compensation.
Addressing "Big Data Killing" and Consumer Deception
The SPC has taken a firm stance against "algorithmic price discrimination," commonly known in China as "big data killing" (dàshùjù shāshú). This occurs when platforms use AI to charge different prices to different users for the same service based on their purchasing history or device type.
The Opinions state that such practices violate the principles of fairness and honesty. If a consumer can prove they were charged an unfair price due to algorithmic manipulation, the court can order the platform to compensate the user. Additionally, the guidelines target deceptive AI livestreaming, where "virtual influencers" are used to sell products without disclosing their non-human nature. The SPC ruled that if AI-generated avatars are used to commit fraud or mislead consumers, the platform and the content creator will be held jointly liable for punitive damages.
Procedural Innovations: AI in the Courtroom
Beyond substantive law, the Opinions introduce procedural rules for the use of AI in litigation. The SPC mandates that if a party uses AI to generate evidence or litigation documents (such as legal briefs), they must disclose this fact to the court.
To prevent "judicial pollution" caused by AI-generated fake evidence, the SPC has authorized the use of blockchain and advanced digital forensics to verify the authenticity of electronic data. Courts are also encouraged to utilize "technical investigators" and expert witnesses to help judges understand the underlying logic of algorithms involved in disputes.
Implications and Future Outlook
The release of these Opinions suggests that China is pursuing a "middle path" in AI regulation—one that is more prescriptive than the United States’ market-led approach but more focused on industrial application than the European Union’s rights-heavy AI Act.
Legal analysts suggest that the SPC’s guidelines will have three primary impacts:
- Judicial Predictability: Companies like Baidu, Alibaba, and Tencent now have a clearer understanding of their liability borders, which may encourage more aggressive R&D in sensitive areas like autonomous driving and medical AI.
- Consumer Confidence: By codifying protections against algorithmic discrimination and deepfakes, the government is attempting to build public trust in digital services.
- Global Influence: As one of the most detailed judicial documents on AI to date, China’s framework may serve as a reference for other jurisdictions, particularly in the Global South, looking to regulate emerging technologies.
The Supreme People’s Court concluded its announcement by emphasizing that these guidelines are not static. As AI technology continues to evolve toward General Artificial Intelligence (AGI), the judiciary will remain flexible, updating its interpretations to ensure that the law continues to serve as both a "propeller" for innovation and a "brake" for ethical risks. The Opinions effectively set the stage for a new era of "intelligent justice," where the rule of law and the power of the algorithm are brought into a delicate, sustainable balance.
Suro Senen
Neng Nana
