An Expert Discussion with Ms Tian-ying Zhao on the Intellectual Property Landscape in China

China has emerged as a crucial battleground for intellectual property rights, reflecting its growing influence in global innovation. The country's commitment to fostering a robust IP environment is evident in the substantial number of patent applications filed annually. In 2023, China issued 920,797 patents, almost three times the 315,245 patents issued by the U.S., which ranked second.[1] In 2024, China remained the top source for Patent Cooperation Treaty (PCT) applications, with 70,160 applications,[2] and had a record high of 20,081 patent applications at the European Patent Office (EPO).[3] As China continues to prioritise IP protection, we spoke with Ms Tian-ying Zhao, partner at Beijing-based IP law firm IntellecPro, to find out more about the latest regulatory changes, notable trends and significant case studies.
 
Have there been any recent regulatory changes or interesting developments?
 
Yes, a significant development took place on 24 April 2025, when China’s Supreme People’s Court and Supreme People’s Procuratorate jointly released a new judicial interpretation titled Interpretation on Several Issues Concerning the Application of Law in Handling Criminal Cases Involving Intellectual Property Infringement. Effective as of 26 April 2025, this interpretation replaced a series of previous judicial interpretations and aligned the application of criminal law with the amendments introduced by the Eleventh Amendment to the Criminal Law enacted in 2021.
 
The new interpretation introduces several key changes that reflect a strengthened commitment to criminal enforcement of IP rights:

• Expanded judicial assistance in evidence gathering in private prosecution cases, requiring courts to assist plaintiffs in obtaining evidence they cannot acquire on their own.
• Increased fines, raising the maximum amount from five times to ten times the illegal gains, thereby strengthening the economic penalties for IP crimes.
• Clarified criteria for determining what constitutes "identical goods or services" in trademark counterfeiting cases, including situations where the goods or services have the same name, or where their functions, purposes, or service content are the same or essentially the same.

 
Importantly, the interpretation also adjusts the criminal thresholds for IP infringement:

• Sets lower thresholds for repeat offenders who commit the same type of IP infringement within two years of a previous criminal or administrative penalty.
• Establishes thresholds for crimes involving counterfeit service marks and online dissemination of protected content.
• Lowers the threshold for patent counterfeiting from RMB 500,000 to RMB 300,000 in direct economic loss caused to right holder.

 
This updated judicial interpretation is expected to bolster the deterrent effect of criminal IP enforcement and provide clearer guidance for prosecution and adjudication.
 
What about notable cases or examples of new case law precedence?
 
One notable example of recent legal precedent in China involves the issue of "reskinned" games – a common practice in the Chinese mobile gaming industry, where developers release new games by retaining the gameplay mechanics of pre-existing games developed by others, while making only superficial changes to the visual and narrative elements.
 
In April 2025, the Supreme People’s Court of China included a key precedent – Chengdu Legou Technology & Shanghai Lilith Network v Shenzhen Jiujiu Mobile & Hainan Tomato (Guangdong High Court, (2023) Yue Min Zhong No. 4326)—in its 2024 collection of model IP cases, shedding light on how courts assess “reskinning” under Chinese copyright and unfair competition laws.
 
In the said case, the plaintiffs developed Rise of Kingdoms, a successful strategy simulation game. The defendants released a game titled Commander on the WeChat mini-program platform, which the plaintiffs argued was a “reskinned” copy of their original game. Despite changes in audiovisual elements, the overall structure, gameplay system, parameters, numerical designs, and even textual errors were found to be nearly identical.
 
The first-instance court found the defendants liable for copyright infringement and unfair competition, ordering cessation of operations and compensation of RMB 10 million in damages plus RMB 500,000 in legal costs. On appeal, the Guangdong High People’s Court upheld the compensation order but reframed the legal reasoning.
 
The appellate court ruled that game rules and gameplay mechanics are not protected expressions under copyright law, nor do they constitute “other intellectual achievements” eligible for copyright-like protection. However, it held that the defendants' conduct– duplicating the entire gameplay design framework and numerical systems while merely replacing graphics – violated the principles of commercial integrity and constituted unfair competition.
 
This case draws a clearer boundary between copyright protection and unfair competition law in the gaming industry. It establishes that while gameplay systems may fall outside the scope of copyright, wholesale duplication of such systems, when combined with only superficial aesthetic changes, can still trigger liability under unfair competition law. The ruling serves as a guiding precedent for future disputes involving game reskinning, and reflects China’s growing emphasis on protecting innovation and fair market order in the digital entertainment sector.
 
Are there any new changes on the horizon and, if so, what implications do you anticipate?
 
Yes, there have been major changes in how trademark cancellation cases are handled in China. Since the second half of 2024, the China National Intellectual Property Administration (CNIPA) has increasingly issued official actions requiring applicants to supplement evidence. These notices demand applicants provide additional proof that the disputed trademark has not been used for three consecutive years. The volume of such notices has grown notably in 2025, especially since April, with many cancellation filings now subject to detailed follow-up requirements.
 
The supplemental requirements have become more diverse and stringent. Applicants are asked to submit:

• Search results from multiple major online platforms and relevant industry-specific websites;
• Basic business information of the registrant, such as business scope, operational status, and trademark filing records;
• New trademark application associated with the disputed trademark and review of refusal submission for the new trademark;
• Where the applicant’s or agent’s address is in the same province as the registrant, on-site investigation report may also be required.

 
Moreover, many applicants are expected to submit a declaration affirming that they have not concealed the identity of the actual applicant or other key facts from the CNIPA.
 
According to the CNIPA, these changes stem from their response to complaints from rights holders, who have reported malicious cancellation filings against their marks (e.g. ungrounded or repetitive cancellations against their registrations). As a result, CNIPA is not only tightening evidentiary standards but also proposing a legislative revision to restrict who may file cancellation requests – from “any entity or individual” to only those with a legitimate interest.
 
Under China’s Trademark Law, any entity or individual may apply to cancel a registered trademark on the ground of non-use. The applicant shoulders limited evidentiary requirements, while the burden of proof traditionally lies with the trademark registrant. To avoid potential retaliation from trademark owners, many non-use cancellations have been filed anonymously through unrelated third parties. Foreign applicants, in particular, often rely on domestic third parties to file on their behalf to avoid the burden of notarising and legalising documents required for filing court appeals.
 
The implications of the recent changes are twofold. The evidentiary burden placed on applicants raises the cost and complexity of pursuing cancellation. For applicants relying on third-party filers to remain anonymous – a common tactic to avoid retaliation or avoid notarisation burdens in litigation – the requirement to disclose the true applicant’s identity may disrupt current practice.
 
How does artificial intelligence affect intellectual property protection?
 
Artificial intelligence is reshaping the landscape of intellectual property (IP) protection in both empowering and challenging ways.
 
On one hand, AI significantly enhances the enforcement and management of IP rights. Advanced AI tools can automatically monitor digital platforms for infringements such as counterfeit goods, pirated content, and trademark misuse. With image recognition and natural language processing, AI enables large-scale, real-time detection that would be unfeasible through manual methods. For example, AI can map out technological landscapes, detect overlapping claims, and assess the relative strength of IP assets based on citation analysis or litigation history. In the trademark domain, AI tools can conduct similarity searches across global databases, flag potentially conflicting marks, and track competitor filings to inform proactive filing strategies.
 
On the other hand, AI poses complex challenges to traditional IP frameworks. A major issue is authorship and inventorship: most jurisdictions require a human creator for copyright or patent protection, creating uncertainty over the status of AI-generated works. Ownership disputes also arise when AI contributes to creative or inventive processes – especially in collaborative settings – making it difficult to determine whether rights should vest in the developer, the user, or the deploying entity. Moreover, the autonomous operation of some AI systems complicates legal liability. If an AI system produces infringing content without human oversight, it becomes unclear who should be held accountable – the developer, the platform, or the end user.
 
Another growing challenge concerns the use of copyrighted content as training data for AI systems. Many generative AI models – whether for text, images, music, or code – are trained on massive datasets scraped from the internet, which often include copyrighted works without the consent or knowledge of the rights holders. This has led to a wave of legal actions against major AI developers, including lawsuits from authors, visual artists, and music publishers. These cases raise unresolved questions about whether such use qualifies as fair use or infringement, and whether new licensing models or legislative frameworks are needed to address the use of protected content in training datasets. The lack of transparency in how training data is collected and used further complicates rights enforcement and accountability.
 
In short, AI enhances the efficiency of IP protection while also exposing gaps in current legal definitions and enforcement mechanisms, calling for new approaches to address these evolving challenges.
 
In light of the recent high-profile cases such as Getty Images v Stability AI and Zhang v Google, what is the stance in your jurisdiction on the use of copyrighted materials in training AI models?
 
In China, current laws do not provide explicit rules on whether the use of copyrighted materials for AI training constitutes infringement. However, recent regulatory developments signal a clear policy direction.
 
The Interim Measures for the Management of Generative Artificial Intelligence Services, issued in July 2023 by the Cyberspace Administration of China and other authorities, explicitly state that pre-training and fine-tuning activities must not infringe others’ lawfully owned IP rights. These measures also encourage the development of public data platforms for AI training. Importantly, the Measures mandate that providers establish complaint and reporting mechanisms, and ensure timely handling and feedback.
 
Additionally, the Basic Security Requirements for Generative Artificial Intelligence Services, released in February 2024, go a step further by requiring providers to identify major IP risks in training data and to avoid using any data that potentially infringing the IP rights of others. These standards also require providers to setup complaint and reporting channels for IP-related issues, and support third-party inquiries into data sources and IP status through such channels.
 
Collectively, these provisions reflect three key principles:
 

• AI training activities must not infringe intellectual property rights, and infringement liability cannot be evaded by claiming 'fair use' as a justification.
• Public data platforms for AI training will be developed, which could help reduce legal risks associated with unauthorised or "wild" data sources, and pave the way for a collective copyright management model to address the complex issue of lawful authorisation for AI training.
• Service providers must establish accessible channels through which copyright holders and third parties can file complaints and inquire about the use of specific content in training datasets. This mechanism is intended to push AI companies to improve the compliance and traceability of their data practices. It not only helps rights holders determine whether their works have been used without authorisation, but also provides a procedural foundation for follow-up actions such as enforcement, regulatory oversight, or legal liability attribution.

 
These regulatory developments demonstrate that while China has yet to establish a comprehensive legal framework specifically governing AI training data, it is actively laying the groundwork for a more structured and rights-conscious environment. By emphasising IP compliance, data transparency, and procedural accountability, Chinese authorities are signalling a firm stance on preventing unauthorised use of copyrighted materials in AI development.
 
How can companies mitigate risks to protect themselves for both the short and long term?
 
IP assets – such as trademarks, patents, designs, copyrights, and trade secrets – are critical to maintaining competitive advantage and market value, but they are also vulnerable to infringement, theft, and regulatory challenges. To mitigate risks and protect themselves in both the short and long term, companies must adopt a proactive and strategic approach to intellectual property (IP) management.
 
In the short term, companies should prioritise identifying and securing their core IP rights through timely registrations in key jurisdictions. This includes conducting clearance searches before launching new products, filing for trademarks and patents as early as possible, and implementing confidentiality agreements to safeguard proprietary information. Regular IP audits can help ensure that the company’s rights are properly recorded, maintained, and enforced.
 
For long-term protection, companies should embed IP considerations into their broader business strategy. This involves actively monitoring the market for potential infringement, developing enforcement protocols, and building an internal culture of IP awareness across departments. As businesses grow and expand internationally, it is essential to understand the IP landscape in different countries and adapt filing and enforcement strategies accordingly. Companies should also consider licensing, joint ventures, and other collaborative arrangements that leverage IP assets while minimizing exposure to legal or commercial risks.
 
Ultimately, treating IP as a strategic asset – rather than a legal formality – helps companies not only protect their innovations and brand reputation but also unlock long-term value and resilience in a competitive global environment.
 
Ms Tian-ying Zhao is a partner at IntellecPro. She has been practicing intellectual property law since 2006, handling both contentious and non-contentious matters. Her expertise particularly lies in trademark portfolio management, trademark oppositions and invalidations, enforcement of trademark and copyright rights, and alternative dispute resolutions. Additionally, she collaborates with foreign associates on patent matters in China. Ms Zhao primarily counsels overseas clients, including many multinational corporations. She is well regarded for her insightful analysis and thoughtful advice on legal strategies. Furthermore, Ms Zhao frequently writes on topics related to trademarks, patents, and copyrights.