Lina Hope
The Supreme People’s Court of the People’s Republic of China has formally released a comprehensive set of judicial guidelines titled "Opinions on Hearing Cases of AI-Related Disputes," marking a pivotal shift in the nation’s legal framework for the digital age. This new judicial interpretation addresses the growing complexities of artificial intelligence (AI) integration across society, providing clear directives on liability for torts, intellectual property (IP) infringement, contract violations, and criminal activities. Beyond substantive law, the SPC has also introduced specialized courtroom procedures and trial mechanisms designed to handle the technical nuances of algorithmic evidence and automated decision-making. These guidelines aim to provide a standardized approach for lower courts as they face a surge in litigation involving generative AI, autonomous systems, and algorithmic bias.
A New Framework for a Rapidly Evolving Technological Landscape
As AI technologies become embedded in everything from financial services to healthcare and creative industries, China’s judiciary has recognized that existing legal statutes often fall short of addressing the unique challenges posed by non-human "decision-makers." The SPC’s Opinions serve as a bridge, translating traditional legal principles into the context of the 21st-century digital economy. The document focuses on four primary pillars: the determination of liability in tort cases, the protection and boundaries of intellectual property, the validity of AI-generated contracts, and the prosecution of AI-assisted crimes.
One of the most significant aspects of the new guidelines is the clarification of "duty of care" for AI developers and service providers. Under the new rules, providers must demonstrate that they have implemented robust safety protocols and transparency measures to mitigate foreseeable risks. If an AI system causes harm—whether through a self-driving car accident or a biased recruitment algorithm—the court will now look at whether the developer adhered to national technical standards and whether the "black box" nature of the algorithm was managed with sufficient oversight.
Chronology of China’s AI Regulatory Evolution
The release of these SPC Opinions is the culmination of nearly a decade of strategic planning and legislative layering. To understand the significance of this document, it is essential to view it within the timeline of China’s broader AI governance strategy:
- July 2017: The State Council issues the "New Generation Artificial Intelligence Development Plan," setting a goal for China to become the world leader in AI by 2030 and emphasizing the need for established laws and ethical norms.
- September 2021: The National Next-Generation AI Governance Professional Committee releases the "Ethical Norms for New Generation Artificial Intelligence," prioritizing human control and privacy.
- January 2022: Regulations on "Algorithmic Recommendations in Information Services" take effect, targeting "filter bubbles" and price discrimination.
- January 2023: The "Provisions on the Administration of Deep Synthesis Internet Information Services" are implemented to regulate deepfakes and synthetic media.
- August 2023: China implements the "Interim Measures for the Management of Generative AI Services," one of the world’s first comprehensive sets of rules specifically for technologies like ChatGPT and Ernie Bot.
- 2024–2025: A series of landmark cases in the Beijing and Guangzhou Internet Courts regarding AI-generated content and data scraping create a demand for high-level judicial uniformity.
- September 2026: The Supreme People’s Court issues the "Opinions on Hearing Cases of AI-Related Disputes" to codify trial practices and resolve discrepancies between regional court rulings.
Intellectual Property and the AI Creative Frontier
A central tension addressed by the SPC involves the copyrightability of AI-generated outputs. Following global debates, the SPC has taken a nuanced stance. While fully autonomous AI creations without human intervention may still lack authorship status, the guidelines provide a path for copyright protection if a human user can demonstrate "significant creative control" through prompt engineering, iterative refinement, and selection.
This builds upon the 2023 Beijing Internet Court ruling in Li v. Liu, where the court recognized copyright in an AI-generated image because the user had provided specific prompts and adjusted parameters to achieve a particular aesthetic result. The SPC Opinions now formalize this logic, instructing judges to evaluate the "human element" in the creative process. Conversely, the guidelines also strengthen protections for original content creators whose works are used to train Large Language Models (LLMs). The SPC specifies that unauthorized scraping of copyrighted data for training purposes may constitute infringement unless it falls under narrowly defined "fair use" exceptions for research or public interest.
Data and Statistics: The Growing Burden on the Judiciary
The necessity for these guidelines is underscored by the rapid increase in AI-related litigation. According to data from the China Academy of Information and Communications Technology (CAICT), disputes involving algorithmic transparency and data privacy rose by 42% between 2024 and 2025.
Furthermore, the Chinese AI market is projected to exceed 600 billion yuan (approximately $83 billion USD) by the end of 2026. With millions of enterprises integrating AI into their workflows, the potential for contractual disputes is immense. The SPC reports that in the first half of 2026, over 15,000 cases were filed involving "smart contracts" or automated service agreements. The new guidelines provide a framework for determining the validity of these contracts, particularly in instances where an AI agent makes a "mistake" that leads to financial loss, assigning liability based on the degree of human oversight and the clarity of the underlying code.
Specialized Procedural Mechanisms and Technical Fact-Finding
Perhaps the most innovative section of the SPC Opinions involves the "Trial Mechanisms" for AI disputes. Recognizing that traditional judges may lack the technical expertise to dissect complex neural networks, the SPC has authorized the expanded use of "Technical Fact-Finders" (技术调查官). These experts will sit alongside judges to provide neutral evaluations of algorithmic behavior.
Key procedural updates include:
- Algorithmic Audits as Evidence: Courts may now compel companies to submit parts of their source code or algorithmic logs for "in-camera" review to determine if bias or negligence occurred.
- Shift in Burden of Proof: In certain tort cases where the plaintiff cannot possibly understand the AI’s internal logic, the burden of proof may shift to the defendant (the developer) to prove that their system was not at fault.
- Digital Evidence Preservation: New rules for the use of blockchain to timestamp and preserve AI-generated evidence, ensuring that digital trails cannot be tampered with before a trial.
Official Responses and Stakeholder Reactions
The release has prompted immediate reactions from China’s legal and tech sectors. A spokesperson for the Supreme People’s Court stated during a press conference that the goal of the Opinions is "to balance the protection of individual rights with the necessity of fostering technological innovation." The SPC emphasized that the law should not be a "stumbling block" for AI development but rather a "guardrail" that ensures the technology remains "controllable and trustworthy."
Leading tech firms, including Baidu, Alibaba, and Tencent, have expressed cautious optimism. In a joint statement via a prominent industry association, these companies noted that "clearer judicial standards reduce the ‘legal uncertainty’ that often chills investment in high-risk, high-reward AI research." However, some legal scholars have raised concerns about the "burden of proof" shift, suggesting it might lead to defensive engineering where companies prioritize legal compliance over system performance.
Professor Zhang Wei, a specialist in digital law at Tsinghua University, remarked, "These guidelines are a significant step toward ‘algorithmic justice.’ By allowing technical experts to participate in the trial process, the SPC is acknowledging that the law must evolve into a multidisciplinary field."
Global Implications and Analysis
The SPC’s move places China at the forefront of global AI jurisprudence, alongside the European Union’s AI Act and various executive orders in the United States. While the EU has focused on a "risk-based" regulatory approach, China’s approach through the SPC is uniquely "remedy-focused," prioritizing how harms are redressed in a court of law.
The implications for international business are profound. Multinational corporations operating in China must now ensure that their AI models comply not only with administrative regulations but also with these specific judicial standards. If an AI system developed in the West is deployed in China and leads to a dispute, it will be judged under these new SPC criteria, particularly regarding data usage and algorithmic transparency.
In conclusion, the "Opinions on Hearing Cases of AI-Related Disputes" represent a landmark effort to bring order to the "Wild West" of the digital frontier. By addressing the granular details of liability, IP, and procedure, the Supreme People’s Court is attempting to create a predictable legal environment that supports China’s ambition to be a global AI powerhouse while maintaining social stability and protecting the rights of its citizens. As these guidelines are applied in courtrooms across the country, they will likely serve as a blueprint for other nations grappling with the same existential legal questions posed by the rise of artificial intelligence.
Asep Darmawan
Suro Senen
