Comparative Legal Perspectives On Copyright Protection For AI-Assisted Content

Author: LawSphereX

Category: Intellectual Property

Published: 2026-08-23T18:51:54.635355+00:00

Updated: 2026-08-26T17:11:45.25+00:00

COMPARATIVE LEGAL PERSPECTIVES ON COPYRIGHT

PROTECTION FOR AI-ASSISTED CONTENT

Agnes Amala Anitha T, Assistant Professor of Law, School of Law for Women, Dhanalakshmi Srinivasan University, Perambalur – 621212 Reshma Ranjith, Assistant Professor of Law, School of Law for Women, Dhanalakshmi Srinivasan University, Perambalur - 621212

ABSTRACT

The rapid evolution of Artificial has transformed the creative landscape, raising unprecedented legal challenges for copyright law. Traditionally, copyright frameworks have been grounded in human authorship, recognizing creativity as an exclusively human endeavour. However, the emergence of AI-assisted tools capable of generating literary, artistic, musical, and digital content has blurred the line between human and machine creativity. This paper examines comparative legal perspectives on copyright protection for AI-assisted works, highlighting historical developments, global approaches, legislative responses, and judicial interpretations across multiple jurisdictions.

While the United States maintains a strict human-centered stance, excluding AI-generated content, the European Union emphasizes liability and substantial human involvement, and China has taken a more progressive approach, extending protection where human contribution is evident. India, in contrast, lags behind with inconsistent recognition of AI’s role and no explicit legal framework. International treaties such as the Berne Convention, TRIPS, and WIPO initiatives remain limited in addressing AIspecific concerns, leaving ambiguity in authorship, ownership, and liability. The findings reveal that the absence of clear, harmonized global standards creates uncertainty for creators, developers, and industries reliant on AI technologies.

This paper argues for the urgent need to reform copyright laws to balance innovation with protection of human creators, while preventing monopolization of intellectual property by corporations through AI. Recommendations emphasize the adoption of transparent data practices, international cooperation, and safeguards to preserve human creativity. Ultimately, a balanced and globally coordinated framework is essential to ensure AI-assisted creativity thrives alongside traditional copyright principles. Keywords: Artificial Intelligence, Copyright Law, Authorship and Ownership, Comparative Legal Analysis, Intellectual Property Rights (IPR)

Introduction Copyright laws delve around human centric works, but we are in the era of Artificial Intelligence. AI Technologies have developed so fast, humans started to adapt to the new technologies for creative works. AI started to dominate the creative industry where humans find a way to make their job easier and opting for shorter time rather than earlier days. AI Assisted works merely assists in the process of creation. The AI based applications can produce contents in various genres, namely literary theatre works, software related creation activities, musical composition and recording works, painting sculptures and other artistic works. Human needs to be innovative according to the current trends they cannot be stuck to the traditional creativity. But in the present scenario, the issue is traditional copyright laws which recognize only on human centric works and human creations.

Earlier we have only one creator called human but the case is different now-days and we have another one called as Artificial intelligence. This creates a legal Ambiguity and uncertainty, to address who should be the creator or author of the AI Assisted content, whether AI or Human or AI developer who trained that Application. If the Copyright Laws are not updated to handle AI assisted content that will lead to chaos and confusion, weakening the rights of human creators. This confusion of AI Assisted content and copy right law will slow down the innovation process in the growth of creative industry. On the other hand, if the laws are clear to include AI’s role and its fair usage it can support human creativity as well as it supports new technologies growth.

Tracing the Historical Development of Copyright Laws The statute of Anne, 1710 was the first copyright law in England, it was a watershed event in the history of copyright, since other laws favoured publishers, it was the only law that favoured authors1. Loi du 19 Juillet, 1793 law of France not treated copyright as economic privilege but as moral right2. Federal Copyright Law,1883 of Switzerland was the first National Law to protect Photographs. Later with an increase in international publications there was a need for international copyright protection. Then arose the Berne Convention, 1886 for protection of literary and Artistic works. Berne Convention mainly laid down principles such as automatic protection, no need of formal registration, National treatment, life plus 50 years rule. More than

180 countries ratified the convention and still acts as the backbone of international copyright

law3. Copyright Act, 1899 of Japan was the earliest to adopt neighbouring rights ahead of many

1 Statute of Anne, 1710, 8 Ann., c. 19 (Gr. Brit.).

2 Loi du 19 Juillet 1793 (Fr.).

3 Berne Convention for the Protection of Literary and Artistic Works, Sept. 9, 1886, 828 U.N.T.S. 221.

countries. Copyright Act, 1957 was the first independent legislation of India Which replaced

1914 Act. Rome Convention, 1961 focused on protecting performers, producers of

phonograms, and broadcasting organizations4. U. S. Copyright Act, 1909 extended the protection over artistic works and raised the term of copyright. Soviet Copyright Law, 1928 of USSR vested copyright to the state, treated creative works belonged to the collective than individual. The Universal Copyright Convention was enacted as an addendum in 1952. The Copyright Act,1976 allowed fair usage of copyrighted works for educational purposes. Trade Related Aspects of Intellectual Property Rights (TRIPS), 1994 expanded coverage to computer programs5. WIPO Copyright Treat (WCT), 1996 and WIPO Performances and Phonograms Treaty (WPPT) started addressing digital era copyright issues6, Marrakesh Treaty, 2013 provided access to visually impaired by balancing accessibility rights. From these early statues Copyright Law has steadily evolved to balance the protection of author’s rights.

Each reform reflects the law’s capacity to adapt to retain its core purpose. 7 Artificial Intelligence’s Impact on Creative Industries Innovation in gradually taking its other form through Artificial Intelligence. AI changes creative industries at high rate, which tends to obscure the distinction between AI assisted works and Human made works. There are many applications such as OpenAI (GPT), Gemini by Google and Stability AI are used in content creation, others like AIVA and MuseNet are composing musical symphonies. The movie industries are getting to harness the technologies of AI in script writing, visual effects, there is Deepfake to improve efficiency in production. DALL- E, DeepDream and Midjourney are excellent in generating digital art and graphic design which creates a huge impact in creative industries.

AI assisted works uses AI to enhance the creative part of the work with an input of human contribution. The large language models or generative AI are being trained by using copyright protected works. AI intentionally or not, it ends up creating a significant number of copyright material distinctly violating the main principles. AI reshapes the creative process by exposing the limitations of copyright regimes. The machine creativity tends to challenge human centric works of copyright law by raising serious questions about authorship, ownership and legal protection in the digital age. 4 Rome Convention for the Protection of Performers, Producers of Phonograms and Broadcasting Organizations, Oct. 26, 1961, 496 U. N. T. S. 43. 5 Agreement on Trade-Related Aspects of Intellectual Property Rights, Apr. 15, 1994, Marrakesh Agreement Establishing the World Trade Organization, Annex 1C, 1869 U.

N. T. S. 299.

6 WIPO Copyright Treaty, Dec. 20, 1996, 36 I.L.M. 65 (1997).

7 Marrakesh Treaty to Facilitate Access to Published Works for Persons Who Are Blind, Visually Impaired or Otherwise Print Disabled, June 27, 2013, 52 I.L.M. 175 (2013).

Global Approaches to AI assisted work and Copyright By revisiting the Berne convention framework, the convention seems to assume that there will be human author for all creative works and therefore lacks a conceptual fit for anything which is termed AI assisted content. By understanding the WIPO’s position on AI and Copyright, WIPO Copyright Treaty, 1996 initially extended its protection to transmitted online works and focuses on safeguarding technological protection and recognizes databases as protected property, but the treaty did not anticipate the rise of AI as a major content creator till date. WIPO is still on the discussion stage on the AI assisted content under the treaty, main debate is whether AI should be treated as tool or Creator.

WIPO’s AI and IP Policy Study, 2019 study investigated the interface of AI with Copyright law and contended that traditional human models of authorship do not fit AI assisted works. WIPO is active in leading global consultations on whether AI assisted works should receive copyright protection or not and also working towards international standards. TRIPS, to be the most comprehensive and extensive international treaty for IPR governance, however, as regards to AI assisted content, the agreement is mute. In 2023, UNESCO adopted first global recommendation on AI ethics which also provided role of AI in creative works. 8 The European Union Artificial Intelligence Act,2024 is the first international legal development that proposes regulation for AI, however it focuses on the transparency and accountability of AI, rather failed to address this issue. In

2019 OECD Principles on AI aimed to link innovation and copyright law in global AI

governance. OECD Principles promoted transparency of AI, explained the need of disclosure of AI training data and discussed that methods to access originality is the need of an hour. OECD also recognizes economic impact of AI in creative industries; human contribution must be remained valued even AI accelerates creativity. These global initiatives reflect their cooperation to ensure that copyright laws remain fostering AI innovation and safeguarding human rights. Comparative Legislative Approaches to AI assisted content and Copyright Law Different countries have different approaches to AI assisted works while some jurisdiction insist only human made work whereas others are moving towards flexible system. The U. S. Copyright Act, 1976 strictly focuses on protecting original works of human authored purely excluded AI work9.

Section 102 and 103 requires human intellectual for upholding originality,

8 UNESCO, Recommendation on the Ethics of Artificial Intelligence (2021),

https://unesdoc.unesco.org/ark:/48223/pf0000381137.

9 Copyright Act of 1976, 17 U.S.C. §§ 101–122 (2018).

while 201 defines creator as the owner. U. S Legislation is rooted towards human cantered approach and currently there is no statute for regulating the current issue. Moving on to the EU AI Act, 2024 regulated AI in general regarding transparency under Article 52, but the law does not address this AI work and copyright issue. Other related directives of EU also protect only human cantered works, while there are exceptions to data training under, Article 3 and 4 of copyright directive allows AI developers to use copyrighted material for data training AI. The copyright, Designs and Patents Act, 1988 be unique in addressing computer generated works under section 9(3)10. UK’s IP office is exploring reforms, expanding from human centric works to AI assisted works.

China’s Legislative stance on AI assisted copyright which reflects a state driven and compliance-oriented model that only aims to prioritize content traceability and developer accountability rather than expanding the definitions of authorship or ownership. India’s copyright law has still not caught up with the challenges of ai- assisted works, where there is no clear rules or no question of authorship which is uncertain and create a grey area11. Overall, most of the countries discussed, approaches towards the issue are fragmented and heavily rooted towards human centric traditional laws. Moving forward, there is essential to accommodate evolving forms of AI assisted creative works While equally maintaining the core principles of traditional Copyright Laws. United States Judicial Approaches Naruto v.

Slatter is significant case which is also known as “Monkey Selfie case” where it’s ruled that copyright law protects only those works created by humans and non – humans cannot acquire rights under Copyright Act12. In Thaler v. Perlmutter court affirmed that copyright protection under the 1976 Act requires human authorship, while AI can be a tool in the creative process but works without any human involvement that is AI generated works are not eligible for protection13. US Copyright Office refused to register Theatre D’opera Spatial, it’s an AI generated artwork by Jason Allen using Midjourney Application, although he argued prompts and edits given by him were creative contribution, it was considered as not eligible for copyright14.

In Zarya of the Dawn US Copyright Office decides contrasting Theatre D’opera Spatial, in this context the author wrote a story including visual layouts by using Midjourney

10 Copyright, Designs and Patents Act 1988, c. 48, § 9(3) (UK).

11 Interim Measures for Generative Artificial Intelligence Services (promulgated by Cyberspace Admin. of China, July 10, 2023, effective Aug. 15, 2023) (China).

12 Naruto v. Slater, 888 F.3d 418 (9th Cir. 2018).

13 Thaler v. Perlmutter, No. 22-1564, 2023 WL 5333236 (D.D.C. Aug. 18, 2023).

14 U.S. Copyright Office, Letter to Jason Allen re: Theatre D’Opera Spatial (Sept. 5, 2022).

Application15. Initially USCO revoked full registration of the work and later copyright was granted only for human authored contribution excluding the part of AI assisted Images. The Kadrey V. Meta Platforms was the first case to address the AI Training copyrighted work and it was set as precedent that AI developers have to get permission from original authors16. There were numerous ongoing cases indicates that judicial approach in AI assisted content and Copyright is still evolving. European Union Judicial Approaches In Infopaq International A/s v. Danske Dagblades Forening, European Court of Justice ruled even tiny snippets that is only 11 words can be protected by copyright under EU Law. This ruling is not about being completely inventive but about author making creative choices. 17 The case Painer v.

Standard established even a portrait photograph can be protected if there is photographer’s creative touch18. This principle is relevant to AI assisted works, meaning if there is substantial human involvement that work could be eligible for copyright protection. The cases YouTube (C-682/18) and Cyando (C – 683/18) clarified that platform providers are also held liable automatically if they publish infringed content. This ruling could extend to AI application using copyrighted works for Data training process. 19 The Ocilion (C – 426/21) case mirrors the same ruling, reaffirms that platform providers are held liable for creating infringing content. Significant decision made in Case No. 10C 13/2023, the Prague Municipal Court AI assisted image with only a prompt without any human involvement cannot be qualified as copyright protected work.

20 This decision leaves a positive opening for protection of AI assisted work if there is a meaningful human creativity in it. Recent rulings in European Union reinforces with significant human input, AI assisted work fall inside the scope of copyright protection. United Kingdom Judicial Approaches UK court held in Nova Productions v. Mazooma Games that player inputs and game visuals generated during gameplay were not copyrightable, as it lacked human creative inputs. 21 Several ongoing decisions in cases like Getty images v. Stability AI, Author’s Guild v. OpenAI

15 Zarya of the Dawn, U.S. Copyright Office Review Board Decision (Feb. 2023).

16 Kadrey v. Meta Platforms, Inc., No. 3:23-cv-03417 (N.D. Cal. filed July 7, 2023).

17 Case C-5/08, Infopaq Int’l A/S v. Danske Dagblades Forening, 2009 E.C.R. I-6569.

18 Case C-145/10, Painer v. Standard VerlagsGmbH, 2011 E.C.R. I-12533.

19 Joined Cases C-682/18 & C-683/18, YouTube & Cyando, EU:C:2021:503.

20 Case C-426/21, Ocilion IPTV Techs. GmbH, EU:C:2023:67.

21 Nova Prods. Ltd. v. Mazooma Games Ltd., [2007] EWCA (Civ) 219.

reflects continuous engagement of judiciary with questions of authorship or copyright ownership and liability in AI assisted work. 22 While the final outcomes are still pending, we could find the trend towards recognition human inputs and stricter scrutiny of AI generated content. China Judicial Approaches China began actively regulating AI assisted content, notably done through “Interim Measures for Generative AI Service” which stresses the need for proper compliance, authenticity. A landmark case, Shenzhen Tencent v. Shanghai Yingxun, recognized AI assisted work and provided copyright protection due to significant human involvement in that work. 23 These ruling signals that AI assisted work gets protection in China if human creativity is evident. The Beijing Intellectual Property Court held in Goa Yang v.

Youku that AI assisted screenshots from video could also be eligible for protection. 24 The case reinforced China’s stance that human involvement is essential for copyright protection. In Li v. Liu, court has recognized AI assisted image since human involvement through prompt, parameter, adjustment and selection of AI services is evident. 25 The Changshu Municipal Court upheld copyright in Lin Chen v. Quin Hong Real Estate Development Co. , Ltd. & Hangzhou Gauss Air Film Technology Co. , Ltd, AI assisted 3D artwork recognized with human authorship where AI was used as creative tool. 26 Yin V. Beijing Intelligent technology Company, the Beijing court’s first AI assisted voice infringement case was held that imitating personal voice via AI without consent violates personal rights27.

In Guangzhou Internet court case, court held that AI service provider is liable for failing to prevent infringing outputs and court advised to used keyword filtering to avoid violations. Overall, China’s approach shows a progressive stance towards AI assisted work. China intents to encourage innovation without diluting human centric traditional laws. India Judicial Approach The Delhi High court held in CBSE V. Aditya Bandopadhyay that a legal entity cannot claim authorship, institutional works without clear individual contribution cannot be qualified. 28 This strongly denotes that India’s authorship principle is Tied to only natural persons. The New York

22 Getty Images (US), Inc. v. Stability AI Ltd., [2023] EWHC (Ch) (UK).

23 Shenzhen Tencent v. Shanghai Yingxun, Beijing Internet Ct. (2019) (China).

24 Gao Yang v. Youku, Beijing IP Ct. (2020) (China).

25 Li v. Liu, Beijing Internet Ct. (2021) (China).

26 Lin Chen v. Quin Hong Real Estate Dev. Co., Ltd., Changshu Mun. Ct. (2021) (China).

27 Yin v. Beijing Intelligent Tech. Co., Beijing Ct. (2022) (China).

28 CBSE v. Aditya Bandopadhyay, (2011) 8 S.C.C. 497 (India).

Times sued OpenAI and Microsoft, alleging that its articles were reproduced without proper authorization. 29 This case highlights the tension in AI data training and Copyright protection, which has not been addressed yet in Indian Law. In India fair usage defence is being used in many of the ongoing cases claimed by AI service provider. In 2020 India’s copyright office made significant history in recognizing AI as co-author. This was the very first time AI was recognized as co-author; however, the decision was revoked because of lack in legal provisions for non- human authors. India’s judicial approach remains largely underdeveloped compared to other jurisdiction, overall, India is lagging behind other nations in accommodating AI and copyright law.

Research Findings Through the study we could find that, there is no existing laws explicitly addressing the AI assisted content under copyright law causing uncertainty and legal ambiguity. Current international frameworks do not adequately cover posed issue. All discussions on AI and copyright are still in the discussion stage, no global frameworks have been adopted yet. Contrasting attitudes towards the issue are adopted across jurisdiction, each taking different stance on AI assisted works. US maintains a strict and conservative stance, refusing to recognize AI generated works and insisting for human authorship. EU enforces strict liability, penalizing even minor infringement, mainly focuses on liability of platform providers.

China provides copyright allows copyright for AI assisted works which has significant human involvement or input, but excludes autonomous AI generated work also strictly prohibits voicebased infringements. In India AI was removed from Co-authorship, showing reluctance to recognize AI as a creator. AI models are trained with copyrighted works which is likely to cause infringement, while human who uses AI automatically becomes the secondary infringer of that particular work. AI generated works also mimics traditional styles like Ghibli art devaluing original artistry and threatening creative livelihoods. AI is not a legal person to grant authorship rights, if rights were given AI created a legal loophole, allowing corporate companies to monopolize creative works and control Intellectual property without any accountability.

Overall, a balanced approach is needed to foster creativity while securing IP rights, ensuring that both human creators and AI contributes fairly recognized and protected Recommendations and Suggestions

1. The study suggests to develop laws addressing AI assisted work and copyright to

29 The New York Times Co. v. OpenAI, Inc., No. 1:23-cv-11195 (S.D.N.Y. filed Dec. 27, 2023).

eliminate legal uncertainty and provide clarity on authorship, ownership and liability of AI in creative works.

2. International bodies should expedite adoption of global frameworks to regulate AI and

copyright, providing clear and unified guidelines that member states can implement, this could regulate cross broader copyright issues.

3. Countries should engage in a collaborative dialogue to harmonize their approaches by

establishing more consistent and equitable legal framework globally.

4. UN and other countries like US, EU should reconsider its strict stance and explore more

flexible approach towards the issue. Legal framework should be updated to address secondary infringement issues.

5. Legal safeguards should be put in place to protect traditional creators are not

undermined by AI Technologies, and frameworks should be designed to ensure AI assisted works benefit human creators and avoid concentration of power within corporations.

6. Greater transparency should be mandated in AI data training, requiring clear consent

from individuals regarding the use of their data, and providing safeguards to protect user’s privacy and control over their personal data. Conclusion The study has explored the intricate intersection of Artificial Intelligence and Copyright Law, emphasizing the evolving challenges and opportunities presented by AI assisted creative works. As AI technologies continue to be advanced, they are reshaping the landscape of creativity and protection of intellectual property. The study highlights the gaps and ambiguities in existing legal framework. The findings from various jurisdictions reveal that there is a need of unified framework. This absence led to uncertainties both for creators and AI developers.

Furthermore, the rapid pace of technological advancement has outstripped the ability of legal systems to keep up with traditional copyright laws, this often ill- suited the realities of AI assisted content. The study also finds that there is need for a balanced approach that fosters creativity while safeguarding the rights of human creators, at the same time, there must be recognition that AI cannot be treated as a legal person, granting Authorship to AI may lead to monopolistic control over creative works, undermining the rights of human creators and smaller entities. Ultimately the study advocates for enhanced transparency in AI data training processes. The recommendation is clear that there is an urgent need for a collaborative, multi-stakeholder approach to create a robust and fair legal framework that can effectively address the challenges

posed by AI in the realm of copyright law.