CN/EN

All
  • All
  • Product Management
  • News and Information
  • Introduction content
  • Business outlets
  • Frequently Asked Questions
  • Corporate Video
  • Corporate Portfolio

On September 7, 2026, the Supreme People’s Court held a press conference on the “Opinions of the Supreme People’s Court on Lawfully Adjudicating Disputes Involving Artificial Intelligence.” Vice President Tao Kaiyuan, Director of the Research Office Zhou Jiahai, Chief Judge of the Third Civil Division Li Jian, and Deputy Director of the Research Office Si Yanli attended the event and answered questions from reporters. The press conference was chaired by Ji Zhongbiao, Spokesperson for the Supreme People’s Court.


Q: We have noticed that, Opinions of the Supreme People’s Court on Lawfully Adjudicating Disputes Involving Artificial Intelligence (hereinafter referred to as the “Opinions”) Based on current pressing and challenging issues, the document provides targeted responses to matters of public concern and key concerns within the artificial intelligence industry. Specifically, how does the “Opinions” appropriately balance the relationship between rights protection and innovative development?


Answer: General Secretary Xi Jinping, in his keynote address at the opening ceremony of the 2026 World Artificial Intelligence Conference and the High-Level Conference on Global AI Governance, emphasized the need to consistently prioritize both development and security, grasp the trends and underlying principles governing AI’s evolution, and continuously refine relevant laws and regulations, policies and institutional frameworks, application standards, and ethical guidelines. This will ensure that AI is safe, reliable, and controllable, enabling it to advance swiftly and steadily as a “thousand-li horse.” The Opinions are firmly anchored in the principle of “balancing development with security,” addressing both the illegal and abusive misuse of AI through legal regulation and strengthening judicial safeguards for citizens’ rights and interests in the digital age, while also unequivocally supporting technological and industrial innovation to provide robust support for the healthy development of AI. Specifically:


With regard to safeguarding the rights and interests of the people, the Opinions set forth the following key provisions: First, they strengthen judicial protection of personality rights. At present, unlawful practices such as AI face-swapping, AI voice imitation, and “online box‑opening” are increasingly prevalent, drawing widespread public concern. In response, the Opinions establish clear rules for holding accountable those who use artificial intelligence to infringe upon rights of portrait, reputation, privacy, and voice, thereby fully leveraging the normative and guiding functions of judicial rulings to promote the ethical and beneficial development of AI. Second, they provide legal protection for personal information rights. Large‑scale, high‑quality training data is the cornerstone of AI technological advancement. To strike a balance between protecting personal information rights and fostering the growth of the AI industry, the Opinions, in accordance with the Civil Code, the Personal Information Protection Law, and other relevant statutes, stipulate that, for the purpose of training AI models, processing personal information that has been lawfully made public within a reasonable scope—and where the individual has not explicitly objected—generally does not constitute an infringement. However, where such processing has a significant impact on an individual’s rights and interests, consent must be obtained in accordance with the law. Third, they regulate, in accordance with the law, practices such as “big data price discrimination” and “impersonating celebrities for product promotion” that harm consumers’ legitimate rights and interests. Recently, “big data price discrimination” driven by AI algorithms has been repeatedly exposed, while “impersonating celebrities for product promotion” has also attracted considerable attention. We cannot expect all consumers to possess discerning judgment; the law must act promptly to safeguard consumers’ lawful rights and interests. To this end, the Opinions clarify that, where an operator uses algorithms to impose unreasonable differential treatment—such as in transaction prices or other terms—on the same goods or services, thereby causing harm to others’ legitimate rights and interests, the people’s courts shall, in accordance with the law, hold the operator liable for corresponding tortious acts. Furthermore, where AI‑driven impersonation of celebrities for product promotion constitutes fraud, and the consumer seeks punitive damages, the people’s courts shall grant such claims in accordance with the law.


With regard to supporting innovation and development, the Opinions set forth the following key provisions: First, they accurately define the principles of liability for torts involving artificial intelligence, ensuring that overly stringent liability regimes are not imposed at an early stage of AI technology and industry development, thereby safeguarding incentives for innovation. The Opinions clarify that, except where the law expressly provides for strict liability or presumed fault, liability arising from infringements of civil rights through the use of AI shall be governed by the principle of fault-based liability. Second, they strictly adhere to legal criteria in defining “AI products” to prevent the overbroad application of product liability. Specifically, the Opinions stipulate that the definition of “product” under the Product Quality Law must be rigorously applied, limiting the scope of AI products to those embodied in physical objects—such as intelligent robots and autonomous vehicles—while excluding AI‑based services that lack a tangible form. Third, they comprehensively and precisely implement a criminal policy that balances leniency with severity, ensuring that criminal cases involving AI are handled in accordance with the law. On the one hand, the Opinions emphasize the need to resolutely impose severe penalties, in accordance with the law, on intentional acts of using AI to endanger national security, infringe upon citizens’ rights, or disrupt social order. On the other hand, they call for prudent and lawful treatment of innovative activities in AI research, development, and application, while upholding the principles of legality and proportionality between crime and punishment. Furthermore, as Vice President Tao previously explained, the Opinions contain specific provisions for the proper adjudication, in accordance with the law, of intellectual property disputes involving AI; these provisions likewise reflect the underlying philosophy of fostering innovation within the framework of the rule of law.


Q: At present, new types of intellectual property disputes arising from AI-generated content are occurring with increasing frequency, drawing widespread attention. Could you please explain what new provisions the “Opinions” introduce regarding the adjudication of IP disputes involving AI-generated content?


Answer: Articles 12 through 16 of the “Opinions” are dedicated provisions addressing intellectual property disputes involving artificial intelligence; in particular, Article 12 provides guidance on the issue of AI-generated content you mentioned. The Outline of the 15th Five-Year Plan explicitly calls for exploring rules to determine the ownership of rights in AI-generated works and to clarify the rights and responsibilities of developers, operators, and users. Furthermore, the State Council’s recently issued “15th Five-Year Plan for Intellectual Property Protection and Utilization” underscores the need to refine the copyright system to better accommodate the development of artificial intelligence. The “Opinions” implement the relevant directives of the CPC Central Committee and promptly address public concerns. Specifically, there are three key aspects:


First, the criteria for determining the liability of relevant parties in cases involving copyright infringement arising from AI-generated content have been clarified. Such infringements implicate multiple stakeholders, including developers, providers, and users of AI technologies. When AI-generated content is alleged to infringe another party’s copyright, the mere fact that the allegedly infringing material was generated by AI cannot serve as a basis for exempting the responsible parties from liability; rather, liability must be aligned with each party’s degree of control and duty of care. In assessing control and duty of care, it is generally necessary to take into account such factors as the technical characteristics and business model of the AI system at issue, the role played by each party in the content-generation process, the source of the training data, the measures taken to mitigate risks, and the extent of any economic benefit derived. On this basis, the law should determine whether the parties bear subjective fault and, if so, the degree of such fault, thereby apportioning liability in a manner consistent with the principles of fairness and reasonableness.


Second, the rules of evidence for disputes involving copyright infringement arising from AI-generated content have been clarified. When a rights holder alleges that an AI service developer or provider has infringed their copyright, they must establish, by way of preliminary evidence, that the infringing content was generated by the AI and that it is substantially similar to the rights holder’s copyrighted work. Given the complexity and opacity of AI technologies, such evidence is not readily accessible to others. Accordingly, if an AI developer raises a non‑infringement defense, they must provide evidence regarding the source of the model’s training data, the training process, the operational mode, and other relevant facts; where necessary, they should also furnish supporting scientific or theoretical grounds. Furthermore, if an AI user knows or ought to know of a prior work and uses AI to generate a work that is substantially similar to that prior work, without any applicable defenses such as fair use, such conduct shall be deemed to constitute infringement.


Third, regulate, in accordance with the law, unfair competition practices involving AI-generated content. Over the past two years, disputes arising from the use of artificial intelligence to engage in unfair competition have been on the rise. Operators exploit AI‑generated text, images, videos, and virtual personas to fabricate traffic and positive reviews; they even employ AI face-swapping to produce fake science‑communication videos for deceptive advertising. Such conduct not only infringes upon the legitimate rights and interests of others but also seriously misleads consumers and disrupts the order of market competition, and should therefore be legally recognized as constituting acts of imitation‑based infringement and false advertising, among others.


It should be noted that, during the drafting and deliberation of the “Opinions,” significant disagreements arose regarding the copyrightability of AI-generated content and the legal characterization of using others’ works without authorization to train large-scale AI models; further clarification is needed. Accordingly, the “Opinions” refrain from addressing these two issues at this stage. At present, artificial intelligence technology is advancing at a rapid pace, giving rise to an ever‑increasing array of new circumstances and challenges. Moving forward, we will continue to closely monitor intellectual property protection in emerging fields such as AI, deepen our understanding of underlying principles through practical experience, systematically distill judicial insights, refine adjudicatory rules, and strive to build broad consensus.


Q: As artificial intelligence accelerates its iterative development, the barriers to infringement posed by technologies such as deepfakes, intelligent data collection, and user profiling have been significantly lowered, posing serious challenges to the protection of personality rights. You just mentioned that the “Opinions” address hot‑button issues involving infringements on personality rights, such as AI‑generated face swaps and voice synthesis, as well as online “box‑opening” practices. Could you elaborate further?


Answer: The right to personality is integral to each individual’s dignity and constitutes the most fundamental and essential right of civil subjects. We emphasize the need to strike a balance between protecting personality rights and fostering the development of the artificial intelligence industry, adopting an approach that is both inclusive and prudent; however, inclusiveness does not equate to indulgence, nor does prudence amount to laissez-faire. In response to public concerns and to guide judicial practice, the Opinions set forth targeted provisions, differentiated according to the specific types of conduct, addressing instances where artificial intelligence technologies are abused or misused to infringe upon personality rights. These provisions primarily include:


First, it is necessary to regulate, in accordance with the law, practices such as “AI face-swapping and voice synthesis” that exploit artificial intelligence to infringe upon rights of portrait, reputation, and voice. In practice, AI‑based deepfakes have drawn the strongest public concern. For example, AI face‑swapping technology can indiscriminately appropriate individuals’ likenesses, making ordinary people potential targets of deepfake manipulation; similarly, AI‑powered voice synthesis allows users to “steal” another person’s voice for just a few yuan, producing results that are convincingly realistic. Recently, the media exposed an illicit supply chain peddling software and tutorials for “one‑click AI cloth removal,” a practice that has sparked widespread public outrage. Article 4 of the Opinions explicitly stipulates: processing a natural person’s name, likeness, or other personal identifiers by artificial intelligence without consent, and then creating, using, or publicly disclosing a virtual digital image that can identify that individual, constitutes an infringement of the right to name, the right of portrait, and other personality rights; likewise, using a natural person’s voice as training data without consent to imitate their timbre, intonation, and pronunciation style, thereby generating a synthetic voice that can be identified, amounts to an infringement of voice‑related rights; and manipulating virtual avatars or synthesized voices to engage in improper conduct, make false statements, or otherwise undermine another person’s social standing likewise constitutes an infringement of the right to reputation, with the perpetrator held liable under the law. Furthermore, in response to the public’s growing concern over the phenomenon of “AI reviving the deceased,” the Opinions include specific provisions to safeguard the personality interests of the deceased in accordance with the law.


Second, it is necessary to regulate, in accordance with the law, the use of artificial intelligence to engage in activities such as “online box‑opening” and “human flesh searches,” which infringe upon others’ right to privacy. “Online box‑opening” and “human flesh searches” are prominent manifestations of online harassment; they not only gravely violate victims’ privacy and seriously disrupt their peace of mind but also undermine the public’s sense of security and pollute the online environment. At present, there have emerged instances in which artificial intelligence is employed to track and analyze publicly available information—such as telephone numbers, online accounts, and social media profiles—of specific individuals, thereby obtaining private data or disclosing and making public such information. In response, Article 5 of the Opinions contains specific provisions addressing these issues.


Third, the rules governing the application of the injunction system for infringement of personality rights have been clarified. Conduct that uses artificial intelligence to infringe upon personality rights is characterized by low costs, rapid dissemination, broad impact, and damage that is difficult to reverse. The traditional post‑facto litigation‑based remedy is inadequate for promptly and effectively safeguarding the legitimate rights and interests of victims; thus, the lawful application of injunctions against infringements of personality rights is particularly important. To this end, Article 8 of the “Opinions” provides explicit guidance. For example, if a victim can demonstrate that they have been subjected to an AI‑generated “pornographic rumor” through face‑swapping, and failure to intervene promptly would result in irreparable harm to their legitimate rights and interests, they may apply to the people’s court for an injunction against such infringement. Upon review, if the conditions are met, the people’s court may, in accordance with the law, order the relevant online user to cease the infringing conduct, or it may likewise order the online service provider or the AI service provider to take timely and necessary measures to prevent further expansion of the harm.


It should be particularly noted that, at present, there are frequent instances of fraud, defamation, slander, infringement of citizens’ personal information, and the production, sale, and dissemination of obscene materials—often carried out using deepfake technologies and other such techniques. Where such conduct constitutes a crime, it must be resolutely and strictly punished in accordance with the law, with criminal liability pursued; the “Opinions” contain specific provisions to this effect.


Q: At present, generative AI software is widely used, and related civil disputes have attracted public attention. I have noted that, with respect to the determination of infringement involving generative AI, the “Opinions” have introduced a “safe harbor” regime. Could you provide an in-depth explanation of this issue?


Answer: Article 1195 of the Civil Code establishes the “notice-and‑takedown rule,” commonly referred to as the “safe harbor rule.” Simply put, when an internet user employs an online service to commit an infringement—for example, posting defamatory remarks about another person—upon discovery, the rights holder is entitled to notify the online service provider to take measures such as blocking or removing the content. Upon receiving a valid notice, the online service provider must promptly implement appropriate measures; otherwise, it shall bear joint liability with the user for any additional damages resulting from its failure to act. Conversely, if the online service provider, upon receiving such notice, promptly takes the necessary steps, it generally will not be held liable for the infringement.


In practice, there is ongoing debate over whether providers of generative artificial intelligence services may invoke the “safe harbor” regime by analogy. After careful examination, we conclude that, although generative AI services differ from traditional online services, they share certain similarities in their underlying operating mechanisms. For instance, both types of services organize vast amounts of information and make it available to users; the outputs generated by large-scale models depend on factors such as the training data, model parameters, and the prompts provided by users. Given that AI service providers cannot foresee in advance the specific content submitted by every user, and considering the sheer volume of generated content, it is impractical to require them to individually predict, review, or block potentially infringing material. Nevertheless, once an AI system produces infringing content—whether due to its own hallucinations or malicious user prompting—and the rights holder issues a notice, the provider is obligated, and well-positioned, to take appropriate measures. Accordingly, there is both a legitimate basis and sound rationale for applying the “safe harbor” regime by analogy to providers of generative AI services.


To this end, the Opinions set forth provisions on the attribution of liability for “AI hallucinations” that infringe upon personality rights, as well as for cases where users maliciously induce AI to violate such rights. Specifically, it clarifies that if content automatically generated by generative artificial intelligence infringes upon personality rights and, upon notification by the rights holder, the AI service provider fails to promptly take necessary measures—such as halting the generation of infringing content—it shall bear tort liability in accordance with the law for the damages caused. Furthermore, if an internet user, by entering infringing prompts or other means, maliciously induces generative AI to produce infringing content, thereby causing harm to others, that user shall bear tort liability under the law. Should the rights holder notify the AI service provider, and the provider fail to promptly adopt necessary measures—such as ceasing the generation of infringing content or blocking relevant generation instructions—and the rights holder seeks civil liability from both the internet user and the AI service provider, the people’s courts shall uphold such claims in accordance with the law. These provisions are consistent with the patterns of technological development and the legislative spirit of the Civil Code, embodying an organic unity of supporting innovation and governance based on the rule of law, and have received unanimous approval during the consultation process.


Q: The impact of artificial intelligence is pervasive across all fields, and judicial proceedings too are confronted with new risks and challenges. For example, some litigation participants submit evidence and other procedural documents generated by AI to the people’s courts without first verifying their accuracy. How does the “Opinion” address such conduct that uses AI to undermine judicial order?


Answer: Litigation participants shall adhere to the principle of good faith in the course of litigation. With the widespread adoption of generative artificial intelligence technologies, it has become increasingly common for litigants to rely on AI‑assisted tools in their proceedings. However, AI is merely an auxiliary instrument; litigants may not invoke “AI hallucinations” or technological neutrality as grounds for absolving themselves of legal liability. Litigation participants are required to carefully scrutinize and verify any evidence and other procedural materials they submit to the court, ensuring their authenticity and accuracy. To guide the public in the responsible use of AI, prevent false materials from entering the litigation process, and safeguard judicial order, the Opinions set forth provisions in three key areas.


First, fraudulent litigation involving artificial intelligence shall be punished in accordance with the law. Where a party exploits the autonomous learning and decision-making capabilities of AI to fabricate factual claims—by means of human intervention or deception, such as deleting or altering identifiers of generated or synthesized content, inputting specific instructions, selectively presenting results, or introducing adversarial interference—the people’s courts shall, in accordance with the law, dismiss the claim and impose fines or detention depending on the severity of the circumstances; if the conduct constitutes a crime, criminal liability shall be pursued in accordance with the law.


Second, conduct lawful regulation of the use of artificial intelligence to fabricate evidence. The Opinions explicitly stipulate that if a party to litigation or any other person uses AI to falsify evidence and thereby obstructs the people’s court from hearing a case, such conduct shall be dealt with in accordance with Article 114 of the Civil Procedure Law—namely, fines or detention may be imposed depending on the severity of the circumstances; if the conduct constitutes a crime, criminal liability shall be pursued in accordance with the law.


Third, the regulations explicitly stipulate that all participants in litigation bear an obligation to verify and clarify AI‑generated documents, case‑search reports, and other such materials. At present, people’s courts have identified numerous instances of false cases generated by artificial intelligence, with the parties and their agents failing to conduct thorough verification prior to submission. In January this year, the media extensively reported on such incidents, and relevant cases have even been included in the people’s courts’ case database. Building on this, the “Opinions” expressly provide that if any document—such as a pleading or a case‑search report—is generated using AI, the party concerned must rigorously verify the authenticity and accuracy of the underlying laws, judicial interpretations, precedents, and other relevant content before submitting it to the court. Furthermore, when filing such materials, they must truthfully disclose any use of AI‑assisted tools and assume legal responsibility for the veracity and accuracy of the information provided. This issue has also drawn significant attention from foreign jurisdictions; we have observed that some countries have issued specific guidelines emphasizing that agents bear the duty to ensure the accuracy of materials submitted to the court.




※ Source: News Bureau of the Supreme People’s Court


Beijing Headquarters Address: No. 8, Jianguomen North Street, Dongcheng District, Beijing 17th Floor, China Resources Building
Contact numbers: 010-64011566, 18811162996

Wuhan Branch Address: Room 1001, 10th Floor, Huangpu International Center, Jiang’an District, Wuhan, Hubei Province

Contact number: 027-82288828

Hefei Branch Address: Rooms 1703–1710, Building E, Weilan Business Harbor, No. 188 Qianshan Road, Shushan District, Hefei, Anhui Province

Contact numbers: 0551-68127192, 15850561916

Related News

CONTACT US

Contact us


Beijing Headquarters

Address: 17th Floor, China Resources Building, No. 8 Jianguomen North Avenue, Dongcheng District, Beijing

Phone: 010-64011566

Email: contact@xinglailaw.com


Wuhan Branch Office

Address: Room 1001, Huangpu International Center, Zhaojiatiao, Jiang'an District, Wuhan City

Phone: 027-82288828

Email: admin@xinglailaw-wuhan.com


Hefei Branch Office

Address:Room 1703-1710, Building E, Weilan Business Port, No. 188 Qianshan Road, Shushan District, Hefei City

Phone: 15850561916

Email: contact@xinglailaw-hefei.com

.

Follow us

.

Digital Star Come

Case Consultation

Experienced lawyers offer free, no-obligation consultations to provide tailored solutions.


%{tishi_zhanwei}%

Copyright 2025 Beijing Xinglai Law Firm

Tags: Website Development:300.cn Beijing Certificate

Business license