Research on the Copyrightability of AI-Generated Results

Jiahe Ye

Houde Academy Dalian University of Technology Dalian City Liaoning Province 116024 China

Keywords: Generative AI, Copyrightability, Originality Standard.

Abstract: The rapid development and widespread application of Generative AI have made the determination of

originality, legal subject qualification, and the creative tool theory. Users can influence AI-generated results

through instructions and parameters, but the primary role of algorithms in production raises questions about

originality. The mainstream view currently holds that AI should not possess legal subject status, although

some scholars propose the possibility of granting it in the future. The creative tool theory posits that AI is a

tool for human creation, but defining user contributions and distinguishing between tools and results remain

challenging. By analyzing relevant domestic and international judicial cases and applying the "minimum

originality" standard, this paper argues for upholding the principle of natural per-sons as authors and using

this standard to differentiate the results of generated works.

1 INTRODUCTION

With the rapid development of artificial intelligence

technology, the application of generative artificial

intelligence in fields such as text, images, music, and

video is becoming increasingly widespread. Taking

ChatGPT as an example, since OpenAI launched

ChatGPT in 2022, generative AI represented by it has

garnered phenomenal attention both in application

and academia (Chen, Jiang, 2024)The results it

generates are highly similar in form to works created

by humans, sparking discussions about whether these

generated results should be protected by copyright

law. The results it generates not only represent a

breakthrough in technical terms but also provoke

profound reflections and discussions in various fields

such as law, ethics, and socio-economics. The

widespread application of generative artificial

intelligence has made the issue of copyright

determination for its generated results increasingly

prominent. On one hand, the content generated by

generative artificial intelligence reflects certain

characteristics of human creation, such as creativity,

aesthetic value, and practicality; on the other hand, its

generation process entirely relies on algorithms, data,

and models, which significantly differs from works

directly created by humans in the traditional sense.

This difference has sparked intense discussions about

whether the results generated by generative artificial

intelligence should be protected by copyright law.

Granting copyright protection could have a

significant impact on the existing copyright legal

system, creative incentive mechanisms, and the

ecology of the cultural industry; conversely,

completely denying copyright protection could stifle

innovation and development in generative artificial

intelligence technology, hindering its application and

promotion in fields such as the cultural and creative

industries. Therefore, conducting in-depth research

on the copyright determination of results generated by

generative artificial intelligence is not only of

significant theoretical value but also holds great

practical significance for balancing technological

development with legal protection and promoting the

healthy development of the cultural industry.

The global academic community has also engaged

in intense discussions regarding the copyrightability

of results produced by artificial intelligence. For

instance, the U.S. Copyright Office, when handling

related cases, insists that content generated by

generative artificial intelligence is not the result of

human creation and therefore should not be protected

by copyright law. This position has received

widespread support in U.S. judicial practice.

European scholars, on the other hand, tend to explore

the copyright issues of results generated by generative

artificial intelligence from the perspective of legal

fiction. They argue that generative artificial

Ye, J.

Research on the Copyrightability of AI-Generated Results.

DOI: 10.5220/0014380700004859

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 1st International Conference on Politics, Law, and Social Science (ICPLSS 2025), pages 353-357

ISBN: 978-989-758-785-6

353

intelligence or its developers and users can be

regarded as legal authors through legal attribution,

thereby granting copyright protection to the generated

results. However, this viewpoint faces numerous

challenges and controversies in practice.

At the same time, in domestic judicial practice, a

series of typical judicial cases, including the case of

Filin Law Firm vs. Baidu, the case of Tencent vs.

Yingxun Technology regarding copyright

infringement, the case of Li vs. Liu regarding the

infringement of work attribution rights and

information network dissemination rights, and the 'AI

Image Infringement Case' in Wuhan High-tech Zone,

showcase the judicial context regarding the

determination of generative AI (Beijing Internet

Court, 2019; People's Court, 2021; Beijing Internet

Court, 2024; Wuhan East Lake High-tech Zone

Court, 2025). They also illustrate the evolution of

China's approach to generative AI from a

conservative judgment based on traditional copyright

law subjects to innovative practical explorations.

This article aims to explore the copyrightability of

results generated by generative artificial intelligence,

analyzing the controversies and challenges related to

originality standards, legal subject qualifications, and

the theory of creative tools. It combines domestic and

international judicial practice cases to propose

reasonable determination standards and suggestions.

Through the review and analysis of existing literature,

this article summarizes the key issues and points of

contention in determining the results generated by

generative artificial intelligence and discusses the

applicability of the 'minimum originality' standard in

this field. Finally, based on a global perspective, it

provides theoretical support and reference for

relevant legal practices through comprehensive

analysis and overview.

2 THEORETICAL OVERVIEW

2.1 Judgment of Originality Standards

Page Setup

Originality is the key standard for determining

whether a work is protected by copyright law. Article

3 of China's Copyright Law states: 'The works

referred to in this law are intellectual achievements in

the fields of literature, art, and science that possess

originality and can be expressed in a certain form

(Standing Committee of the National People's

Congress, 2020). Therefore, the results generated by

generative artificial intelligence must meet the

requirements of originality to constitute a work.

The judgment of originality for results generated

by generative artificial intelligence is subject to much

controversy. Zhu Ge believes that the originality of

AI-generated content should depend on the

intellectual input of the user during the generation

process. By inputting prompts, setting parameters,

and other means to design and adjust the generated

content, users reflect personalized choices and

judgments. This intellectual input endows the

generated content with originality, which should be

protected by copyright law (Zhu, 2024). On the other

hand, Wang Qian emphasizes that the judgment of

originality for AI-generated content should be based

on the autonomy of the algorithms. He argues that AI-

generated content is an application of algorithms,

rules, and templates, lacking the personalized choices

and judgments of human creators, and therefore does

not meet the requirements of originality under

highlights the actual participation and intellectual

input of users in the AI creation process, which not

only aligns with the realities of technological

development but also better accommodates the

widespread application of AI in the creative field. In

contrast, while Wang Qian's viewpoint is logically

clearer and emphasizes the requirements of copyright

law for creative subjects, it may be overly strict in

practical application. Completely excluding

overlook the actual contributions of users in the

creative process, resulting in some original AI-

generated content failing to receive the legal

protection it deserves. Furthermore, as technology

advances, the quality and complexity of AI-generated

content continue to improve, and the influence of

users on the generated content is also increasing. This

makes Zhu Ge's viewpoint more flexible and

adaptable in the current domestic judicial practice,

while Professor Wang Qian's viewpoint has not yet

gained widespread recognition in domestic practice.

2.2 Legal Subject Qualification

Regarding whether generative artificial intelligence

possesses legal subject qualification, there are

differing opinions in academia. The mainstream view

holds that humans are always the legal subjects, while

another view suggests that artificial intelligence may

potentially constitute a legal subject. For instance, Shi

Yongjing believes that with the continuous

development of artificial intelligence technology, it

may be possible to grant legal subject qualification to

artificial intelligence in the future, allowing it to enjoy

rights such as copyright (Shi, 2019). However, this

ICPLSS 2025 - International Conference on Politics, Law, and Social Science

354

viewpoint has not yet gained widespread acceptance.

Nonetheless, it is undeniable that under the wave of

generative artificial intelligence development, the

discussion of legal subject qualification will be

further explored in the future, and this viewpoint

holds significant thought-provoking value. However,

does this contradict the spirit of human subjectivity?

Is it a confusion regarding the subjects of legal

protection? Could it potentially enter judicial

practice? These are among the many legal issues that

this viewpoint needs to address, and it is evident that

the current level of productivity cannot yet support it.

Professor Wang Qian believes that copyright law

aims to protect human creative achievements, and as

a technological tool, artificial intelligence cannot

understand or utilize the incentive mechanisms of

legal subject qualification (Wang, 2023).

Furthermore, even if artificial intelligence possesses

high intelligence and autonomy in the future, whether

it should be granted legal subject qualification still

requires careful consideration to avoid disrupting the

existing legal system and social order.

2.3 The Theory of Creative Tools

Cui Guobin supports the viewpoint of the theory of

creative tools, arguing that generative artificial

intelligence is a tool for human creation, and the

results it generates should be regarded as works

created by humans using the tool. He suggests that

while users cannot fully predict the generated results

when using generative artificial intelligence for

creation, the prompts and parameters they input to

some extent determine the direction and style of the

generated content, and therefore should be seen as

contributing to the creation of the work (Cui, 2024).

The author believes that this viewpoint aligns with the

development and demand for new productive forces

in the era of artificial intelligence, encouraging

human use and creation of new tools, and reflecting

the innovative spirit promoted by copyright law. It is

a theoretical basis that is more in line with the existing

favorable for the development of new productive

forces. However, this viewpoint faces practical

challenges in defining user contributions and

distinguishing between creative tools and generated

results. Therefore, there is an urgent need to clarify a

clear determination standard through the reference

and exploration of judicial practices both

domestically and internationally.

3 EXPLORATION OF JUDICIAL

PATHWAYS FOR COPYRIGHT

OF AI-GENERATED WORKS

3.1 The Cautious Attitude of Early

Cases

Before 2023, courts generally adhered to the

traditional copyright law principle of 'natural person

creation' as the core, denying the work attributes of

AI-generated content and strictly upholding the

'natural person creation' principle. In the 2019 case of

Filin Law Firm vs. Baidu, the court ruled that content

generated by computer software intelligence does not

constitute a work in the sense of copyright law. The

legal logic behind this is that current laws stipulate

that works should be completed by natural persons,

and AI-generated content does not reflect the original

expression of a natural person; the actions of software

developers and users do not fall under creative acts.

The court emphasized that although AI-generated

content possesses a certain degree of originality, any

deviation from the 'natural person creation' principle

should be approached with caution and should protect

relevant intellectual contributions within the existing

legal framework (Beijing Internet Court, 2019). In

2020, the case of Tencent vs. Yingxun marked a new

development, where the court recognized AI-

generated articles as works. In this case, Tencent

generated articles through its developed intelligent

writing system, and the court found that this process

reflected human intellectual investment in data

selection, judgment, and analysis, meeting the

requirements of copyright law for works. This ruling

indicates that the court began to pay attention to the

human intellectual contributions behind AI-generated

content and recognized its work attributes under

specific circumstances (People's Court, 2021).

Overall, early legal determinations regarding the

cautious, emphasizing the 'natural person creation'

principle while gradually exploring the originality of

AI-generated content and the human intellectual

contributions behind it.

3.2 The Emergence of Landmark Cases

After 2023, judicial practice began to recognize the

copyrightability of AI-generated works under

specific conditions, with landmark cases including:

the Beijing Internet Court's 'AI Text-to-Image' case,

where the court explicitly stated that if AI-generated

images reflect the user's 'personalized expression'

Research on the Copyrightability of AI-Generated Results

355

(such as inputting prompt words and adjusting

parameters), they meet the originality requirements of

works and are protected by copyright law (Beijing

Internet Court, 2024). This case provides an

important precedent for subsequent rulings. The

Wuhan East Lake High-tech Zone Court's 'AI Image

Infringement Case' (2025): The plaintiff, Wang,

generated images by adjusting keywords and style

parameters and completed copyright registration. The

court recognized that his creative process reflected

conceptualization and aesthetic choices, deeming the

generated work an intellectual achievement, and

ordered the defendant to pay 4,000 yuan in damages

(Wuhan East Lake High-tech Zone Court, 2025). This

case further refined the 'human participation'

standard, emphasizing the user's control and

foresight. Notably, the Wuhan court adopted a

'minimum originality' standard in this case,

downplaying the examination of creative motivation

and focusing more on the distinctiveness of the

results. The author believe this determination method

can serve as an important reference for contemporary

judicial and legislative practices. Additionally, this

ruling is conducive to stimulating the widespread

application and innovative development of artificial

intelligence generation technology in the field of

artistic creation. It also clarifies that China's copyright

law protects original expressions rather than ideas or

concepts themselves.

3.3 Global Perspectives and

Experiences

The determination of copyright for content generated

by generative artificial intelligence has also formed

several controversies internationally. Among them,

the United States has consistently maintained that

the "Copyright Registration Guide: Works

Containing AI-Generated Materials" was published,

stating that the author must be a natural person, and

the term "author" in copyright law does not include

non-human entities such as artificial intelligence. If a

work is entirely generated by artificial intelligence

without any human creative input, it is not protected

by copyright. When a work contains elements of

human creativity, such as creative selection and

arrangement of AI-generated materials, or

modifications that meet copyright protection

standards, making the work as a whole reflect the

creative control of a human author, copyright

protection can be applied for. Copyright only protects

the aspects of the work that are created by humans

landmark case Stephen Thaler v. Shira Perlmutter et

al., the court held that human authorship is a

fundamental requirement for copyright protection,

and the authors referred to in copyright law only mean

humans; non-human authors are not the objects of

creative incentives under U.S. copyright law (Stephen

Thaler v. Shira Perlmutter et al, 2023). Furthermore,

the notion that authors must be natural persons is

affirmed by international copyright conventions. The

moral rights of authors stipulated in Article 6bis of

the Berne Convention, as well as the rights of authors

after death as provided in Article 7bis, are universally

recognized as the legal basis that authors can only be

natural persons, which has continuously influenced

subsequent global copyright legislation (Xiong,

Zhang, 2024). Additionally, the European Union has

adopted a comprehensive and flexible legislative and

policy framework regarding copyright protection for

generative artificial intelligence outputs. Through the

AI Act and copyright directives, the EU seeks to find

a balance between protecting creators' rights and

promoting technological innovation. Purely AI-

generated content is generally not considered to meet

the conditions for copyright protection, as only

natural persons can be regarded as authors. However,

for AI-assisted content, if it demonstrates creative

human input, it may obtain copyright protection

(Jones, 2025). Countries around the world reflect a

cautious yet innovative attitude towards the

challenges of copyright for artificial intelligence-

generated works.

4 CONCLUSION

In summary, the determination of the results

generated by generative artificial intelligence is a

complex and multidimensional issue that requires a

comprehensive consideration of legal, technical,

ethical, and socio-economic aspects. The author

believes that, based on a review of domestic and

international literature, the determination of

generative artificial intelligence should still adhere to

the principle that natural persons are the authors, and

the "minimum originality" standard can be used to

distinguish the results of generated works. Future

research can further explore the specific applications

and legal issues of generative artificial intelligence

results in different fields and scenarios, as well as

how to promote the healthy development of

generative artificial intelligence technology through

institutional design and policy guidance. At the same

time, it is necessary to strengthen international

ICPLSS 2025 - International Conference on Politics, Law, and Social Science

356

communication and cooperation, learn from

advanced experiences and practices abroad, and

jointly promote the resolution of issues related to the

determination of generative artificial intelligence

results.

REFERENCES

Beijing Internet Court. 2019. Beijing Filin Law Firm v.

Beijing Baidu Network Technology Co., Ltd. Dispute

over Infringement of Right of Attribution, Right to

Protect the Integrity of Works, and Right of Information

Network Dissemination.

Beijing Internet Court. 2024. Li XX v. Liu XX Dispute over

Infringement of Right of Attribution and Right of

Information Network Dissemination - AI-Generated

Image Copyright Infringement Case.

https://elksslpkulaw.aa.sjuku.top/pfnl/08df102e7c10f2

06e0dc41964639771d8919f255da684f3bbdfb.html?ke

yword=%E4%BA%BA%E5%B7%A5%E6%99%BA

%E8%83%BD&way=listView

Chen, R., Jiang, Y. 2024. Research on the Governance of

Generative AI: Taking ChatGPT as an Example.

Studies in Science of Science 42(1):21-30.

Cui, G. 2024. The Originality Contribution of Users in AI-

Generated Works. China Copyright (3):24-29.

Jones, D. 2025. EU Issues Report on Relationship Between

Generative AI and Copyright.

People's Court. 2021. Tencent v. Yingxun Technology

Dispute over Infringement of Copyright.

https://elksslpkulaw.aa.sjuku.top/pfnl/c05aeed05a57db

0af99000eb86b762d22b982782a8882dc3bdfb.html?ke

yword=%E4%BA%BA%E5%B7%A5%E6%99%BA

%E8%83%BD&way=listView

Artificial Intelligence. Journal of Social Sciences of

Jilin University 59(4): 88-95+221.

Standing Committee of the National People's Congress.

2020. The Copyright Law of the People's Republic of

China (Article 3). Beijing: Law Press.

Stephen Thaler v. Shira Perlmutter et al. 2023. Case 1:22-

cv-01564-BAH (D.D.C. August 18).

Guidance: Works Containing Material Generated by

Artificial Intelligence. 88 FR 16190, 16191, 16192.

Wang, Q. 2023. Re-examining the characterization of AI-

generated content in copyright law. Political and Legal

Forum 41(4):3-16.

Wuhan East Lake High-tech Zone Court. 2025. "AI

Generated Image Infringement" copyright dispute case.

https://mp.weixin.qq.com/s/XGS-

JDGXgVqsw9oXPhpoaA

Xiong, Q., Zhang, W. 2024. Global Trends and Local Paths

in the Regulation of Copyright for AI-Generated

Content. Intellectual Property (11):59-76.

Zhu, G. 2024. Is AI-generated content (AIGC) protected by

Research on the Copyrightability of AI-Generated Results

357