131 Reshaping Sri Lankan Intellectual Property Law: The Influence of Artificial Intelligence Sandali Fonseka1* 1Legal Apprentice (Chambers of Kuvera de Zoysa, PC ) Corresponding author *: shamalkifonseka@gmail.com Abstract As Artificial Intelligence (AI) systems increasingly perform autonomous creative and inventive functions, the foundational concepts of authorship and inventorship in intellectual property (IP) law are being challenged. This Article explores the complex intersection between AI and intellectual property (IP) rights. Additionally, the article examines the impact of AI on copyright and patent regulations, specifically within the context of the Sri Lankan legal framework. It critically examines the challenges surrounding authorship and ownership of AI-generated works. The central research questions address whether AI-generated works can qualify for authorship and receive copyright protection; how inventorship is determined in AI-generated inventions; whether non-human creators can be legally recognized; the adequacy of the existing IP regime in Sri Lanka; and the reforms required to build a future-ready, balanced legal framework. A doctrinal and comparative approach was employed to analyse gaps in statutes and trends in judicial decisions across various jurisdictions. Therefore, the study highlights legal and ethical concerns arising from AI-driven innovation by analysing relevant case law, international legal instruments, and evolving policy frameworks. The results indicate an urgent requirement to revise definitions related to authorship, originality, and inventive step to incorporate contributions from non-human entities. The study concludes by suggesting a legal framework that is prepared for the future, aiming to harmonize innovation, ethical considerations, and the rights of human creators in the era of digital advancement. The paper ultimately proposes strategies to balance protecting intellectual property and fostering innovation in the era of artificial intelligence. Keywords: Artificial intelligence; Copyrigh; Intellectual property;Patent 1. Introduction The emergence of artificial intelligence (AI), particularly generative AI, marks a pivotal phase in technological advancement, reshaping innovation, creativity, and the application of intellectual https://doi.org/10.54389/ZATJ4908 132 property (IP) law. While AI is not a novel concept, its rapid evolution and widespread adoption have significantly blurred the lines between human and machine-generated creativity. As AI systems now independently produce text, music, images, and software code, traditional notions of authorship, ownership, and infringement are being fundamentally challenged. This paper explores whether existing IP laws, particularly in the domains of copyright and patent, are sufficiently equipped to govern AI-generated content. It identifies gaps in the current legal frameworks and assesses whether new, future-ready policies are necessary to address these challenges1. In the context of balancing innovation with protection, the study emphasizes the need for policymakers, legal scholars, and technologists to understand the legal and ethical implications of AI-generated works. By examining key case law, comparative international approaches, and recent policy developments, the research offers a critical evaluation of how IP law must adapt to remain relevant in the age of AI. Ultimately, the paper contributes to the broader academic discourse by offering strategic recommendations for harmonizing intellectual property protection with the need to foster AI-driven creativity and innovation. It aims to lay a foundational understanding for ongoing legal reform and informed policymaking in this rapidly evolving field. 2. AI-Generated Works and The Ownership The rise of generative AI has introduced complex questions regarding the ownership and authorship of AI-generated content such as art, music, and literature. These works, created through machine learning and neural networks, blur traditional boundaries of creativity and challenge existing legal frameworks that are grounded in human authorship. Under Sri Lanka’s Intellectual Property Act No. 36 of 2003, copyright protection is explicitly tied to human creators, with Sections 6, 10, and 11 recognizing authorship and moral rights only in relation to natural persons.Consequently, AI-generated works fall outside the scope of protection, leaving creators, developers, and users without legal certainty regarding ownership or exploitation rights.This legal gap raises significant risks for industries increasingly relying on AI, as questions of infringement, licensing, and economic value remain unresolved. Addressing these challenges requires not only doctrinal clarity but also consideration of comparative models where jurisdictions are experimenting with concepts such as joint ownership or sui generis protections for AI-generated works. AI-generated works have garnered significant attention due to their capacity to produce original and innovative creations. These outputs embody a distinctive fusion of human-designed 1 Alexander Cuntz, Carsten Fink, Hansueli Stamm, “Artificial Intelligence and Intellectual Property: An Economic Perspective”(World Intellectual Property Organization,2024) < https://www.wipo.int/edocs/pubdocs/en/wipo-pub-econstat-wp-77-en-artificial-intelligence-and- intellectual-property-an-economic-perspective.pdf > accessed 14th May 2025. 133 algorithms and autonomous machine creativity, blurring the traditional boundaries of authorship and challenging conventional concepts of creativity. One of the major challenges in determining ownership of AI-generated works lies in assessing the extent of human involvement and contribution2 .While AI systems can autonomously produce creative outputs, they rely heavily on human programmers, curated datasets, and algorithmic training. This raises a fundamental legal question: Who holds rightful ownership of an AI-generated work? Is it the AI system itself, the human who programmed it, the organization that owns the AI, or a combination of these parties? Furthermore, can an AI legally be recognized as an “author” or “inventor” under current intellectual property laws?3 Sri Lanka’s Intellectual Property Act No. 36 of 2003 limits copyright protection to works produced by natural persons. The provisions on economic and moral rights (Sections 6, 10, and 11) firmly ground authorship in human identity, thereby excluding AI-generated content from protection. Since AI lacks legal personhood and creative intention, current Sri Lankan law does not provide copyright protection for AI-generated works. Attributing ownership of computer-generated works presents a complex and layered challenge, deeply rooted in legal interpretation and statutory definitions. Determining ownership is further complicated by the roles of programmers, dataset curators, and AI operators. Questions arise over whether rights belong to the AI’s developer, user, or the entity that owns the system4. The Act provides limited guidance in this regard, although case law, such as The Commercial Bank of Ceylon v. Director General of Customs[2003] 2 Sri L.R5, demonstrates the relevance of licensing and contractual arrangements in resolving ownership disputes involving software6. Recent legal developments have sparked debate over the rightful ownership of AI-generated works. Courts and scholars have considered various options, including granting rights to the AI’s developer, the user, or though controversially, to the AI system itself. Notably, the Thaler v. Comptroller-General of Patents case (DABUS7) in the UK clarified that while AI cannot be recognized as an inventor, the owner of the AI system may assert inventorship based on control and 2 White C, Matulionyte R., “Artificial Intelligence Painting The Bigger Picture For Copyright Ownership.”,(SSRN,2019) < https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3498673 > accessed 14th May 2025. 3 Kalin Hristov, “Artificial Intelligence and the Copyright Dilemma” (2017), vol 57, IDEA: The IP Law Review accessed 15th May 2025. 4 Kaushalya Wickramanayake, “Factum Special Perspective: Redefining Copyright Laws – Addressing AI- Generated Works and Authorship Rights in Sri Lanka”(Factum,2024)< https://factum.lk/tech- cooperation/factum-special-perspective-redefining-copyright-laws-addressing-ai-generated-works- and-authorship-rights-in-sri-lanka/ > accessed 15th May 2025. 5 The Commercial Bank of Ceylon v. Director General of Customs[2003] 2 Sri L.R 6Sri Lanka Law Reports ,”The Commercial Bank of Ceylon v. Director General of Customs” accessed 5th Sep 2025. 7 Thaler v Comptroller-General of Patents, Designs and Trade Marks 2023 UKSC 49. 134 ownership. Similarly, in Nova Productions v. Mazooma Games (UK) 8and Tencent v. Shanghai Yingxun Technology Co. Ltd (China)9, the courts attributed copyright to software developers, emphasizing their substantial role in determining the output. These decisions underline a trend toward recognizing human or corporate agents who influence the AI process, although they also reveal limitations in current legal doctrines. Moreover, in MGM Studios v. Grokster10, the U.S. Supreme Court demonstrated how copyright and patent principles intersect, suggesting that insights from one IP field may influence the other. Collectively, these rulings highlight the legal system’s evolving approach to AI-generated content and signal the need for more nuanced frameworks that reflect the complexities of non-human creativity. Internationally, common law jurisdictions such as the U.S., U.K., and Australia require human authorship for copyright protection. The Naruto v. Slater case in the U.S.11, where a monkey could not claim authorship for a photo it captured, illustrates courts' reluctance to recognize non-human entities as authors. Similarly, AI-generated content is not eligible for copyright in the absence of significant human creative input. However,in the case of Sony Corp. v. Universal City Studios12, redirects attention from human authorship to the broader objective of fostering public benefit and creativity. From this perspective, AI-generated creations could be eligible for copyright protection because they contribute to promoting innovation and the development of useful arts, irrespective of the lack of a human author. Proposals have been made for new legal frameworks, such as recognizing “AI-assisted authorship” or establishing joint ownership between humans and AI. Comparative models like France’s sui generis software protection framework also highlight potential paths forward. As AI systems become central to creative industries, the need for updated legislation is urgent. Sri Lanka must consider reforms, including statutory definitions that address AI-generated content and clearer licensing norms, to ensure its IP regime remains future-proof and innovation-friendly13. 3. AI and Copyright Law A central concern in modern intellectual property law is whether works produced by generative AI qualify for copyright protection. Traditionally, copyright law links authorship to human creativity and consciousness. Since AI lacks autonomy and self-awareness, it is generally viewed as a tool, not a creator. As IBM (2023) notes, AI should be considered a system that assists humans rather than independently invents. Unlike software like Photoshop, which requires human input, generative AI creates content with minimal human involvement, complicating the attribution of authorship. Moreover, AI systems often rely on vast datasets, sometimes containing copyrighted 8 Nova Productions Ltd v Mazooma Games Ltd & Ors (CA)[2007] EWCA Civ 219. 9 Tencent Company v Yingxun Company, Case No. Y0305MC No. 14010. 10 MGM Studios, Inc. v. Grokster, Ltd., 545 U.S. 913 [2005]. 11 888 F.3d 418 (9th Cir. 2018). 12 Sony Corp. of America v. Universal City Studios, Inc., 464 U.S. 417 (1984). 13 Rita Matulionyte, ‘Copyright in AI-generated works: Lessons from recent developments in patent law’ (2022)19(2)Scriped file:///C:/Users/Sandali%20Fonseka/Downloads/ssrn-3974280.pdf Accessed 17th May 2025. 135 material, to produce new outputs. This practice raises serious legal concerns regarding infringement, especially when the content is used commercially without consent. In jurisdictions like the United States, Section 107 of the Copyright Act of 1976 outlines the fair use doctrine, permitting certain uses such as text and data mining (TDM). However, its application to generative AI remains legally unsettled. Ongoing cases, such as Getty Images v. Stability AI (UK) 14and Andersen v. Stability AI Inc. (US)15, address unauthorized use of copyrighted materials in AI training and question whether fair use can apply. Sri Lanka’s Intellectual Property Act No. 36 of 2003 currently lacks provisions specific to AI-generated works. Section 5 of the Act defines an author as a “physical person,” thereby excluding non-human creators. Legal scholars, such as Sidath Gajanayaka, suggest reconsidering authorship models to include joint authorship between the human prompter and the AI system16. Additionally, Sri Lanka’s copyright law does not explicitly include a fair use doctrine, unlike the U.S. system. This legal gap limits adaptability to evolving technologies and highlights the need for reforms. Such reforms should include licensing structures for AI developers, statutory fair use provisions, and clear guidelines on the status of AI-generated content17. 4. AI and Patent Law The rise of artificial intelligence has introduced new complexities to the concept of inventorship under patent law. AI systems, capable of producing novel and economically valuable inventions, challenge the traditional requirement that a human inventor must be identified to assert patent rights. A central issue is whether AI-generated inventions can be patented and, if so, who should be recognized as the inventor. While AI cannot yet operate fully autonomously, it plays a critical role in the invention process. Legal systems worldwide, however, continue to insist that inventors must be natural persons. For example, in Commissioner of Patents v. Thaler [2022], the Full Court of Australia ruled that only a human can be listed as an inventor under current law. Similar decisions were reached in Thaler v. Commissioner of Patents [2023] (New Zealand) and by the UK Supreme Court in 2023 . Despite this, South Africa became the first country to grant a patent naming an AI (DABUS) as the inventor. While this decision marked a global first, its legal basis remains uncertain under South African law. Contrastingly, an Australian court in 2021 temporarily accepted DABUS as an inventor, highlighting international inconsistencies and the absence of a unified approach. The traditional notion of “inventiveness” has always been rooted in human traits like creativity and 14 Getty Images v Stability AI [2025] EWCA Civ 749. 15 Andersen et al v. Stability AI Ltd. et al, No. 3:2023cv00201. 16 Apsara Rodrigo,”Sri Lanka’s copyright laws need to be updated for artificial intelligence: Lawyers”( 2024) < https://economynext.com/sri-lankas-copyright-laws-need-to-be-updated-for- artificial-intelligence-lawyers-150509/ > Accessed 5th September 2025. 17 Chaga Bihari Mahingoda, “Intellectual Property Rights in the Era of Artificial Intelligence: Navigating the Challenges and Expanding the Boundaries”,2023, SLIIT International Conference on Advancements in Sciences and Humanities,< https://rda.sliit.lk/bitstream/123456789/3622/1/235 245%20Intellectual%20property.pdf > accessed 14th May 2025. 136 imagination. However, as AI systems become more sophisticated, they increasingly meet the substantive criteria for patentability, novelty, inventive step, and industrial applicability, even if no human is directly involved. Importantly, patent laws generally assess whether an invention meets these criteria, rather than who invents it. Legal scholars suggest that where AI is involved, ownership should be addressed through contractual agreements between stakeholders, developers, data providers, users, and funders18. Others propose requiring at least minimal human involvement to satisfy legal formalities. In summary, while global patent systems are beginning to confront the challenges posed by AI, most jurisdictions still require human inventorship. Without reforms, this limitation could undermine innovation. A harmonized legal framework is needed to provide clarity on the treatment of AI- generated inventions and to balance innovation incentives with legal certainty19. 5. AI and Traditional Intellectual Property Fields The integration of artificial intelligence (AI) into traditional intellectual property (IP) domains, particularly patents and copyrights, has significantly disrupted longstanding legal concepts. This section analyses how AI-driven innovation challenges and reshapes the conventional framework, emphasizing the urgent need for reform. Impact on Patents-AI technologies, such as machine learning and advanced algorithms, are increasingly enhancing the invention process and are even being used in patent examination systems. Despite these benefits, AI complicates traditional legal standards, especially in determining inventive step, novelty, and the level of human contribution. Patents offer inventors exclusive rights for a fixed term, typically 20 years to protect and commercialize their inventions. In Sri Lanka, Section 62(1) of the Intellectual Property Act defines an invention as a practical solution to a technological problem, covering both product and process innovations. To qualify, inventions must meet three criteria outlined in Section 63: novelty, inventive step, and industrial applicability. Sri Lanka follows an absolute novelty standard, requiring global uniqueness. Although Sri Lankan law does not categorically exclude software or AI-generated inventions from patent eligibility, it lacks clear provisions on inventiveness attributed to AI. Sections 62(3) and 66 touch on limitations based on morality, public safety, and environmental risks, which could be interpreted to restrict AI-generated inventions. Section 67 assigns patent rights to the inventor without specifying whether this must be a human, though the use of gender-specific pronouns in Section 70(1) implies a natural person. The ambiguity extends to ownership structures. Section 69(2) implies that inventions made using employer resources belong to the employer, suggesting that if AI were considered a tool, its owner could claim rights over its 18 Gloria J. Miller, “Stakeholder roles in artificial intelligence projects”(2022) < https://www.sciencedirect.com/science/article/pii/S266672152200028X > 19 Apsara Rodrigo,’ Sri Lanka’s copyright laws need to be updated for artificial intelligence: Lawyers’( Economynext,2024) < https://economynext.com/sri-lankas-copyright-laws-need-to-be- updated-for-artificial-intelligence-lawyers-150509/> accessed 17th May 2025. 137 outputs. This interpretation parallels practices in other jurisdictions but lacks statutory clarity in Sri Lanka. Impact on Copyright-AI also challenges foundational principles of copyright law, which assumes a human author. Section 6 of Sri Lanka’s IP Act requires originality and a human creator, while moral rights (Sections 10 and 11) are tied to human dignity and identity. Consequently, AI- generated works fall outside current copyright protections. The absence of statutory recognition for AI contributions necessitates a legal rethink. AI’s increasing role in content creation, literature, music, and visual art raises critical questions about authorship, ownership, and the adequacy of existing legal definitions. Copyright protection for AI-generated works remains legally ambiguous. As such, Sri Lanka’s legal system must consider updates to accommodate human-AI collaboration and joint authorship models. Fair Use and Reverse Engineering-Sri Lanka’s IP framework also lacks a robust fair use doctrine, unlike the United States, which provides greater flexibility for innovative uses like text and data mining. The growth of AI-generated content and its reliance on vast datasets make fair use provisions critical. Additionally, practices like reverse engineering, relevant to AI and software innovation, lack clear legal treatment in Sri Lanka. While some jurisdictions permit it under fair use, especially for interoperability and innovation, Sri Lanka has not defined its stance, creating uncertainty for developers and tech companies. 6. Conclusion The challenges posed by AI to Sri Lanka’s intellectual property regime cannot be resolved through broad reform alone; they demand targeted and actionable measures. First, the Intellectual Property Act No. 36 of 2003 should be amended to explicitly define authorship and inventorship in a technology-neutral manner. For example, provisions could recognize “AI- assisted authorship,” where rights are shared between the human prompter and the AI system’s owner, thus ensuring legal clarity in creative industries. Second, the introduction of a statutory licensing framework for AI developers and users would help manage ownership disputes and promote responsible use of training datasets. Third, a narrowly tailored fair use clause, particularly for text and data mining, should be incorporated to support research, innovation, and interoperability while safeguarding copyright holders’ interests. Fourth, in the patent domain, Sri Lanka should adopt a hybrid model that requires minimal human involvement but allows recognition of AI-generated inventions through contractual allocation of rights among developers, users, and employers. Finally, judicial and regulatory bodies should receive specialized training on emerging technologies, enabling consistent interpretation and enforcement of IP law in AI-related disputes. By pursuing these concrete reforms, legislative clarity, licensing mechanisms, fair use provisions, contractual certainty, and capacity-building, Sri Lanka can create an IP framework that balances 138 innovation with protection. Such measures would not only address current gaps but also position the country to adapt effectively to future technological disruptions. References Primary Sources 1. 888 F.3d 418 (9th Cir. 2018). 2. Getty Images v Stability AI [2025] EWCA Civ 749. 3. MGM Studios Inc v Grokster Ltd 545 US 913 (2005). 4. Nova Productions Ltd v Mazooma Games Ltd & Ors [2007] EWCA Civ 219. 5. Sony Corp of America v Universal City Studios Inc 464 US 417 (1984). 6. Tencent Company v Yingxun Company, Case No. Y0305MC No. 14010. 7. Thaler v Comptroller-General of Patents, Designs and Trade Marks [2023] UKSC 49. 8. The Commercial Bank of Ceylon v Director General of Customs [2003] 2 Sri LR Secondary Sources 9. Andersen et al, v. Stability AI Ltd. et al, No. 3:2023cv00201. 10. Chaga Bihari Mahingoda , ‘Intellectual Property Rights in the Era of Artificial Intelligence: Navigating the Challenges and Expanding the Boundaries’ (2023) SLIIT International Conference on Advancements in Sciences and Humanities https://rda.sliit.lk/bitstream/123456789/3622/1/235%20245%20Intellectual%20property.pd f accessed 14 May 2025. 11. Cuntz A, Fink C and Stamm H, ‘Artificial Intelligence and Intellectual Property: An Economic Perspective’ (World Intellectual Property Organization, 2024) https://www.wipo.int/edocs/pubdocs/en/wipo-pub-econstat-wp-77-en-artificial- intelligence-and-intellectual-property-an-economic-perspective.pdf accessed 14 May 2025. 12. Hristov K, ‘Artificial Intelligence and the Copyright Dilemma’ (2017) 57 IDEA: The IP Law Review https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2976428 accessed 15 May 2025. 13. Matulionyte R, ‘Copyright in AI-generated Works: Lessons from Recent Developments in Patent Law’ (2022) 19(2) Scriped file:///C:/Users/Sandali%20Fonseka/Downloads/ssrn- 3974280.pdf accessed 17 May 2025. 14. Miller GJ, ‘Stakeholder Roles in Artificial Intelligence Projects’ (2022) https://www.sciencedirect.com/science/article/pii/S266672152200028X accessed 14 May 2025. 15. Rodrigo A, ‘Sri Lanka’s Copyright Laws Need to Be Updated for Artificial Intelligence: Lawyers’ (EconomyNext, 2024) https://economynext.com/sri-lankas-copyright-laws- 139 need-to-be-updated-for-artificial-intelligence-lawyers-150509/ accessed 5 September 2025. 16. White C and Matulionyte R, ‘Artificial Intelligence Painting the Bigger Picture for Copyright Ownership’ (SSRN, 2019) https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3498673 accessed 14 May 2025. 17. Wickramanayake K, ‘Factum Special Perspective: Redefining Copyright Laws – Addressing AI-Generated Works and Authorship Rights in Sri Lanka’ (Factum, 2024) https://factum.lk/tech-cooperation/factum-special-perspective-redefining-copyright- laws-addressing-ai-generated-works-and-authorship-rights-in-sri-lanka/ accessed 15 May 2025.