Do users infringe rights—do AIGC platforms have an "immunity pass"?
TechFlow Selected TechFlow Selected
Do users infringe rights—do AIGC platforms have an "immunity pass"?
Can technology truly be neutral?
Navigating Web3 tides with focused insightsAuthor: Xiao Sa Team
Recently, the first case involving copyright infringement by an AIGC service provider was adjudicated, sparking intense discussion about legal compliance issues surrounding AIGC technology. Taking typical text-to-image AIGC platforms currently available on the market as an example, when users generate and disseminate content that may infringe upon others' copyrights, are these platforms truly powerless? Today, Sa Jie's team discusses with you the copyright infringement liabilities and risk mitigation strategies for text-to-image AIGC platforms.
01 User’s Post-Generation Infringement — Is the Platform Jointly Liable?
Typically, text-to-image AIGC platforms operate at the application layer (consumer-facing) and lack the capability to develop large models themselves. Instead, they rely on foundational and intermediate technologies developed by others, offering services that convert textual descriptions into images based on user needs. Therefore, how AIGC outputs are used and distributed depends largely on individual users, making it difficult for platforms to anticipate or monitor subsequent usage patterns. When AIGC-generated content is found to be substantially similar to existing copyrighted works, rights holders may argue that both the platform and the user are jointly liable for infringement, potentially subjecting the platform to unforeseen liability.
Whether an AIGC platform should bear responsibility for a user's subsequent infringing activities remains controversial. The key issue lies in determining whether the platform had subjective fault—a critical element in establishing liability. According to the Summary of Guiding Principles from the Intellectual Property Tribunal of the Supreme People's Court (2022):
From the perspective of subjective fault, joint infringement typically involves three scenarios: First, acts carried out with shared intent—this constitutes classic joint infringement. Second, acts resulting from shared negligence, where collective carelessness or overconfidence causes harm. Third, combined acts involving both intentional and negligent behavior, where multiple parties with differing levels of fault contribute through interconnected actions that result in damage.
In other words, for a platform to be held liable, it must have at least been negligent. Regarding the determination of negligence, "objective negligence" is a widely accepted doctrine in tort law across jurisdictions. This standard evaluates conduct against the level of care expected of a reasonable person in society. If a party fails to meet this duty—violating societal expectations and trust—and does not take reasonable preventive measures against foreseeable harms, they are deemed negligent unless a legally recognized exemption applies.
As seen in China’s first AIGC service provider infringement case, the court emphasized that compensation liability hinges on the defendant’s fault. While AIGC tools possess certain instrumental characteristics, service providers must still exercise reasonable care when delivering such services. The involved platform failed to fulfill its duty of reasonable care and thus was found subjectively at fault, warranting partial liability. Conversely, if an AIGC platform actively implements compliance measures and fulfills its duty of care, even if users later engage in infringing activities, the platform cannot be held liable due to lacking causation (i.e., absence of the subjective fault element required for infringement).
02 Applicability of the Technological Neutrality Principle to Text-to-Image AIGC Platforms
The digital era has brought the concept of "technological neutrality" into sharp focus, with safe harbor rules serving as one of its primary manifestations. In copyright law, this principle is also known as the "substantial non-infringing uses doctrine," originating from the landmark U.S. Sony Corp v. Universal City Studios case:
In that case, the defendant Sony sold video recorders capable of recording television programs either in real time or via timers. Users could also pause and fast-forward to skip advertisements. The plaintiff argued that although Sony did not directly infringe copyrights, its product facilitated, encouraged, or materially contributed to third-party infringement, constituting indirect infringement. However, the majority opinion held that once Sony sold the devices, it lost control over how consumers used them. Without the ability or authority to supervise downstream use, Sony should not be held indirectly liable. To justify this conclusion, the court referenced the "staple article of commerce" doctrine from patent law and introduced the principle of technological neutrality into copyright jurisprudence: so long as a technology or product has substantial legitimate, non-infringing applications—even if it can also be misused for infringement—the provider cannot be held liable merely because infringement occurs.
At its core, the technological neutrality principle limits secondary liability within reasonable and acceptable bounds, reflecting a dynamic balance between fostering innovation and protecting intellectual property. Correspondingly, China's intellectual property legal framework embodies this spirit, particularly through the "notice-and-takedown" mechanism. Under China’s Civil Code, when online users commit infringement via network services, rightsholders may notify the service provider to remove, block, or otherwise address the infringing content. Upon receiving such notice, if the provider promptly removes the content, it may avoid liability.
Sa Jie’s team believes that while technological neutrality serves as a useful framework for allocating responsibility, its legitimacy and applicability must be grounded in established tort principles. Rather than debating whether AIGC platforms qualify as Internet Service Providers (ISPs) or Internet Content Providers (ICPs), or whether they automatically benefit from notice-and-takedown protections, we should return to the fundamental elements of tort liability. Case-by-case analysis based on evidence and facts should determine whether an AIGC platform breached its duty of care and thus bears subjective fault.
03 How Can Text-to-Image AIGC Platforms Fulfill Their Duty of Reasonable Care?
As previously noted, most text-to-image AIGC platforms do not possess the technical capacity to develop or train large models independently. Instead, they integrate third-party APIs to deliver services. During operation, platforms must manage risks associated with both external developers and end-users. With regard to third-party developers, AIGC platforms can conduct comprehensive assessments of their data collection practices and algorithmic ethics, requiring full disclosure of data sources, processing methods, algorithm mechanisms, and security oversight. Additionally, clear contractual agreements should define rights, responsibilities, and liability allocation to mitigate technological risks arising from third parties.
To manage risks related to users’ post-generation activities, platforms can implement compliance measures in accordance with the Interim Measures for the Administration of Generative Artificial Intelligence Services and the Regulations on the Administration of Deep Synthesis Technology for Internet Information Services:
① Provide a prominent labeling function and prompt users to apply visible identifiers; for services capable of generating or significantly altering content, ensure that AIGC outputs are clearly labeled in a reasonable location;
② Establish a sensitive keyword database, refine inclusion criteria and procedures, and employ technical or manual review to screen user inputs and generated content, blocking inappropriate material promptly;
③ Formulate transparent management rules and platform guidelines, incorporating clauses regarding ownership of generated content, liability for infringement, and usage restrictions into service agreements and user terms, ensuring that risks and obligations are communicated clearly, conspicuously, and reasonably to guide users in understanding and responsibly using AIGC technology;
④ Establish efficient and accessible complaint and reporting channels, publicly disclose handling procedures and response timelines, and respond promptly to reports with timely feedback;
⑤ Upon discovering illegal content, immediately implement corrective actions such as halting generation, stopping transmission, or removing content. If a user’s subsequent conduct is suspected of infringement, the platform should issue warnings, restrict functionality, suspend, or terminate service according to law and contract, preserve relevant records, and report to competent authorities.
04 Final Thoughts
The principle of technological neutrality plays a vital role in balancing technological innovation and copyright protection. Compliance is not merely a corporate “safeguard” in competitive markets—it is a strategic advantage. Sa Jie’s team recommends that AIGC enterprises collaborate closely with legal and compliance teams to build robust, efficient compliance systems that secure sustainable competitive edges. Should your AIGC business require legal support, feel free to contact Sa Jie’s team!
That concludes today’s discussion. Thank you, readers!
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
