ABSTRACT
The rapid advancement of Artificial Intelligence (AI) has introduced complex legal challenges, particularly in the realm of intellectual property rights. As AI systems evolve from being mere tools to autonomous inventors capable of generating novel and useful inventions without direct human intervention, questions arise about the applicability of traditional patent law frameworks—most of which are grounded in human-centric notions of creativity and inventions. This research explores the patentability of AI-generated inventions in the context of South Asia, focusing on four key jurisdictions: India, Pakistan, Bangladesh, and Sri Lanka.
Through doctrinal and comparative legal analysis, the study examines the extent to which the current patent laws in these countries can accommodate inventions generated independently by AI. It interrogates the statutory definitions of “inventor” and assesses whether AI systems can meet the criteria of novelty, inventive step, and industrial applicability under existing legal standards. The paper also considers recent international legal developments, such as the DABUS litigation in the United States, United Kingdom, Australia, and South Africa, to contextualize the global debate and extract relevant lessons for South Asian policymakers.
The research identifies significant gaps in South Asian IP laws, including the absence of explicit provisions for non-human inventor-ship, institutional limitations of patent offices, and a lack of regional coherence. In light of these challenges, the paper proposes a set of legal and policy recommendations aimed at harmonizing regional patent laws. These include adopting a co-inventor-ship model, developing sui generis frameworks for AI-generated works, and fostering collaboration among South Asian countries through platforms like SAARC or BIMSTEC.
The paper underscores the urgent need for South Asian legal systems to modernize their IP regimes in response to technological realities, thereby ensuring innovation is both protected and promoted in a fair, forward-looking manner.
INTRODUCTION
The intersection of artificial intelligence (AI) and intellectual property (IP) law represents one of the most complex and evolving frontiers in legal scholarship. As AI systems become increasingly sophisticated—capable not only of performing tasks but also of generating inventions without direct human intervention—the foundational assumptions of patent law are being put to the test. Traditionally, patent systems across the world have been premised on the idea that inventor-ship is a uniquely human endeavor. This belief underpins key legal criteria such as originality, inventive step, and industrial applicability. However, with the emergence of AI programs that can autonomously devise solutions, design chemical compounds, or even develops mechanical innovations, the legal community is compelled to reconsider the boundaries of patentability.
In recent years, high-profile cases such as those involving the DABUS AI system have reignited global debate over whether a non-human entity can be recognized as an inventor under current patent law frameworks. Jurisdictions such as the United Kingdom, the United States, and the European Union have largely denied this recognition, citing statutory requirements that link inventor-ship to natural persons. Nonetheless, the divergent outcomes in other countries, including South Africa and Australia, reveal a growing tension within international IP regimes and the need for legal systems to adapt to technological advancements.
While much of this discourse has taken place in Western legal systems, South Asian countries now face similar challenges. India, Pakistan, Bangladesh, and Sri Lanka are gradually becoming hubs for AI development, yet their IP laws remain rooted in colonial-era statutes or early post-independence legal frameworks that do not anticipate the complexities introduced by AI. For instance, the definition of “inventor” remains ambiguous or explicitly tied to human agency, creating uncertainty around the ability to protect of AI-generated works. Moreover, regional patent office’s lack the institutional preparedness to assess and process such applications, further compounding the problem.
Despite their shared historical and legal foundations, South Asian countries exhibit differing degrees of progress in modernizing their IP regimes. This lack of harmonization is problematic, especially in a globalized innovation ecosystem where legal certainty and crossborder recognition of patent rights are crucial. Without a unified regional approach, these countries risk either stifling innovation by excluding AI-generated inventions or undermining the integrity of their patent systems by inconsistently interpreting statutory provisions.
This research sets out to examine the legal and policy challenges related to the patentability of AI-generated inventions in South Asia. It seeks to assess whether the existing patent law frameworks in these jurisdictions are equipped to handle the realities of machine-generated innovation and to what extent legal reforms are necessary. By drawing upon comparative jurisprudence and regional analysis, the study aims to contribute to a more cohesive understanding of how South Asia can position itself in the evolving global IP landscape, where AI is increasingly becoming a key driver of technological progress.
RESEARCH METHODOLOGY
This study employs a doctrinal and comparative legal research methodology to explore the challenges surrounding the patentability of AI-generated inventions within South Asian jurisdictions. The research primarily focuses on a qualitative analysis of existing legal frameworks, case law, and scholarly literature to evaluate how current patent laws in India, Pakistan, Bangladesh, and Sri Lanka address or fail to address the issue of AI as an inventor.
The doctrinal approach involves an in-depth examination of primary legal sources such as statutes, regulations, and judicial decisions related to patent law in the selected countries. These sources include landmark legislation like the Patents Act, 1970 of India, and corresponding statutes in neighboring jurisdictions. International legal instruments and guidelines issued by institutions like the World Intellectual Property Organization (WIPO) also form a critical part of the analysis, providing a global context for the study.
Complementing this, the comparative method analyzes how different jurisdictions interpret and apply the concept of inventor-ship, particularly in relation to AI-generated innovations. This involves a review of significant cases such as the DABUS litigation in countries like the United Kingdom, the United States, Australia, and South Africa. By comparing these legal responses with those of South Asian countries, the research identifies gaps, inconsistencies, and potential models for reform.
Additionally, the study incorporates an analytical perspective to assess the adaptability of existing legal principles—such as novelty, inventive step, and industrial applicability—to AI inventions. It examines the underlying policy considerations and technological realities influencing legislative and judicial attitudes.
The research consciously limits itself to doctrinal and qualitative analysis, without engaging in empirical data collection. Its goal is to contribute to scholarly discourse by offering recommendations aimed at legal harmonization and reform in the South Asian intellectual property landscape.
NATURE OF AI-GENERATED INVENTIONS AND THE QUESTION OF INVENTORSHIP
Artificial Intelligence (AI) is reshaping the way innovations are created. Traditionally, inventions have been the product of human intellect, involving creativity and problemsolving. However, with the advancement of AI, machines are now capable of independently generating new inventions without direct human input. This development challenges the conventional legal frameworks around patentability, which are built on the premise that inventions are the result of human creativity.[1]
AI-generated inventions are those created autonomously by computer programs or algorithms that can analyze data, identify patterns, and develop novel solutions without human guidance. This is distinct from AI-assisted inventions, where human inventors use AI as a tool to aid their creative process. The emergence of AI systems capable of producing inventions on their own raises critical questions about the concept of inventor-ship under patent law.
The patent system requires an inventor to be a natural person—someone who intellectually conceives the invention. This principle is deeply embedded in patent statutes worldwide, including those of South Asian countries like India, Pakistan, Bangladesh, and Sri Lanka, where the law explicitly or implicitly limits inventor-ship to humans. When an AI system creates an invention independently, the legal system struggles to identify who should be recognized as the inventor, or whether such inventions can be patented at all.
Internationally, this issue has led to significant legal debates. For example, the DABUS AI system was named as an inventor in several patent applications across different jurisdictions.
However, patent authorities in places like the UK and the US rejected these applications on the grounds that inventors must be human. Conversely, courts in countries such as Australia and South Africa have been more receptive to rethinking traditional notions of inventor-ship, although no global consensus has yet been reached[2].
Beyond inventor-ship, AI-generated inventions also challenge patentability criteria like novelty and inventive step, which rely on human intellectual contribution. Additionally, questions arise about the ownership of patents for AI inventions—whether rights should belong to the AI’s creator, user, or remain unassigned.
In South Asia, these challenges are especially pressing since the region’s patent laws have not evolved to address AI inventions explicitly. This gap could hinder innovation and create inconsistencies in patent protection. There is a clear need for legal reforms that acknowledge the unique nature of AI-generated inventions and provide clarity on inventor-ship and ownership to keep pace with technological progress.
INTERNATIONAL DEVELOPMENTS AND JUDICIAL TRENDS
The issue of AI-generated inventions and their patentability has become a focal point in global intellectual property discourse, with several jurisdictions grappling with how to integrate AI into their existing legal frameworks. This has led to landmark judicial decisions and policy debates that significantly influence how countries might approach the question of AI inventor-ship and patent protection.[3]
One of the most prominent international cases is the series of patent applications involving the AI system known as DABUS (Device for Autonomous Bootstrapping of Unified Sentience). The inventor applications submitted in countries such as the United States, United
Kingdom, European Union, Australia, and South Africa named DABUS itself as the inventor
rather than a human. This unprecedented move directly challenged the traditional patent law principle that an inventor must be a natural person.[4]
In the United States, the United States Patent and Trademark Office (USPTO) rejected the DABUS application on the grounds that patent law requires a human inventor. The USPTO’s position was upheld on appeal, emphasizing the statutory language that inventor-ship is limited to individuals. Similarly, the UK Intellectual Property Office and courts dismissed the application, reaffirming that current law does not permit non-human inventors.
In contrast, Australia’s Federal Court demonstrated a more flexible approach. Initially, the Australian Patent Office refused the application, but the Federal Court later ruled that the law did not explicitly prohibit an AI system from being named an inventor.. South Africa also showed some openness by allowing AI to be recognized as an inventor, marking it as one of the few jurisdictions willing to adapt traditional concepts.[5]
European patent authorities and courts have generally maintained a conservative stance similar to that of the US and UK, stressing that patent laws were drafted with human inventors in mind and that legislative amendments would be required to accommodate AI inventor-ship.
These divergent responses highlight the lack of global consensus on how to treat AIgenerated inventions within the patent system. They also underscore the need for clear, forward-looking legal frameworks that can balance protecting innovation with maintaining the integrity of patent systems.
For South Asian countries, observing these international trends is vital. They provide valuable lessons and cautionary tales for harmonizing local laws with global standards. The willingness of courts in Australia and South Africa to rethink traditional inventor-ship models signals the possibility of reform, whereas the resistance in the US and Europe emphasizes the challenges ahead. In a region still developing its AI ecosystem, South Asia can benefit from these comparative insights to craft responsive and harmonized patent policies that promote innovation while ensuring legal clarity.
CHALLENGES IN HARMONIZING SOUTH ASIAN IP LAWS ON AIGENERATED INVENTIONS
South Asia, comprising countries such as India, Pakistan, Bangladesh, and Sri Lanka, is witnessing rapid technological growth, especially in the fields of artificial intelligence and innovation. However, the legal frameworks governing intellectual property rights in these countries largely remain rooted in traditional concepts developed long before AI’s rise. This creates substantial challenges for harmonizing patent laws across the region to effectively address AI-generated inventions.[6]
One of the primary challenges is the foundational definition of “inventor” in patent laws. Most South Asian patent statutes explicitly or implicitly restrict inventor-ship to natural persons, reflecting a legal tradition where creativity and invention are inherently human traits. This limitation poses a problem when inventions originate autonomously from AI systems, as these frameworks do not recognize non-human inventors, leaving AI-generated inventions in a legal grey area. The absence of clear guidelines on inventor-ship leads to uncertainty about whether such inventions are patentable at all and who should hold the rights—the AI developer, user, or another entity.[7]
Another significant challenge lies in the divergence of patent office practices and judicial interpretations within the region. While countries like India have made some strides in updating their patent regulations, others still rely heavily on outdated statutes and lack institutional preparedness to evaluate AI-related inventions. This disparity hinders regional cooperation and creates a fragmented patent environment that can discourage cross-border innovation and investment.7
The technological complexity of AI-generated inventions further complicates the application of traditional patentability criteria such as novelty, inventive step, and industrial applicability. Evaluating whether an AI invention meets these standards requires a nuanced understanding of AI processes, which many patent examiners and judges in South Asia may currently lack.
Without specialized training or updated examination guidelines, patent office’s risk inconsistent or flawed decisions, undermining the credibility of the patent system.[8]
Moreover, the absence of regional harmonization mechanisms specific to AI inventions adds another layer of difficulty. South Asia’s existing intellectual property cooperation frameworks, such as those under SAARC, have yet to prioritize AI-related legal reforms. This gap limits the ability of member states to share best practices, develop common guidelines, or coordinate on legislative updates that could create a more unified approach.
Lastly, policy and ethical considerations around AI inventor-ship and ownership add complexity to legal reforms. Debates about whether AI should be granted legal personhood or sui generis rights, or how to balance incentives for human inventors with recognition of AI’s role, remain unsettled. Policymakers in South Asia must navigate these broader questions while crafting laws that encourage innovation without compromising existing IP principles.
RECOMMENDATIONS FOR LEGAL REFORM AND IN SOUTH ASIA
Given the rapid advancement of artificial intelligence and its growing role in creating inventions, South Asian countries face an urgent need to reform their intellectual property laws to address AI-generated inventions effectively. Legal systems in India, Pakistan, Bangladesh, and Sri Lanka must evolve to accommodate the unique challenges posed by AI, while fostering innovation and maintaining the integrity of patent regimes. The following recommendations outline key steps for reform and regional harmonization.[9]
Firstly, South Asian patent laws should explicitly redefine the concept of “inventor” to clarify whether AI systems can be recognized as inventors or whether a different legal framework is necessary. This could involve introducing sui generis rights for AI-generated inventions or allowing human inventors associated with AI development to claim inventorship under specified conditions. Clear legislative provisions will provide certainty for inventors, patent offices, and courts, reducing litigation and inconsistency.9
Secondly, capacity building within patent offices is critical. Examiners and adjudicators should receive specialized training on AI technologies to better understand the nature of AI-
generated inventions and apply patentability criteria appropriately. Establishing dedicated AI units within patent offices could streamline application processing and ensure consistency across jurisdictions.
Thirdly, regional cooperation through platforms like SAARC should be strengthened to promote the harmonization of patent laws related to AI inventions. Collaborative efforts can facilitate the development of common guidelines, information sharing, and coordinated policy initiatives. A harmonized regional framework would benefit inventors and investors by providing a predictable and efficient system across South Asia, encouraging cross-border innovation.[10]
Fourthly, South Asian countries should engage with international bodies such as the World Intellectual Property Organization (WIPO) to align national reforms with evolving global standards. Participation in international dialogues on AI and IP can help South Asia anticipate future developments and incorporate best practices.
Additionally, ethical and policy considerations should guide reforms. Legislators must carefully balance incentivizing human creativity while recognizing AI’s increasing role in innovation. This includes addressing ownership rights, liability issues, and the broader societal impact of AI inventions. Public consultations involving stakeholders from academia, industry, and civil society can ensure that reforms are inclusive and forward-looking.
CONCLUSION AND FUTURE OUTLOOK
The emergence of artificial intelligence as a creator of inventions marks a profound shift in the landscape of intellectual property law. This transformation challenges long-standing legal principles, particularly the concept of inventorship and patentability criteria, which were originally designed with human inventors in mind. South Asia, as a rapidly developing region with burgeoning AI capabilities, stands at a crossroads where it must reconcile traditional patent frameworks with the innovative realities introduced by AI-generated inventions.
This research highlights the significant challenges faced by South Asian countries—including India, Pakistan, Bangladesh, and Sri Lanka—in addressing AI inventorship under existing patent laws. Current legislation in these jurisdictions restricts the definition of an inventor to
natural persons, leaving AI-generated inventions in a legal limbo. This gap creates uncertainty around patentability, ownership rights, and the scope of protection for AI-driven innovations. Without timely reforms, these ambiguities risk undermining the region’s potential to foster technological progress and compete on the global stage.
International judicial decisions, such as those related to the DABUS AI system, underscore the global nature of this challenge. While some jurisdictions like Australia and South Africa have shown willingness to reconsider traditional inventorship concepts, many others, including the US and UK, maintain strict human-centric interpretations. These divergent approaches provide both cautionary lessons and models for potential reform in South Asia. They emphasize the importance of legislative clarity, institutional preparedness, and alignment with emerging global norms.
The path forward for South Asia involves a combination of legal reform, capacity building, and regional cooperation. Revising patent laws to explicitly address AI-generated inventions—whether through expanding the definition of inventor-ship or introducing sui generis rights—will provide much-needed legal certainty. Enhancing the expertise of patent examiners and judges in AI technologies will improve the quality and consistency of patent administration. Additionally, fostering collaboration among South Asian countries can facilitate harmonized approaches that benefit inventors and innovators across borders.
Looking ahead, it is crucial that policymakers in South Asia engage with international intellectual property bodies and participate actively in global discussions on AI and patent law. This engagement will help the region stay abreast of technological advancements and best practices. Furthermore, addressing ethical, ownership, and policy questions related to AI inventions will be essential to crafting balanced laws that protect innovation while promoting social welfare. The incorporation of AI-generated inventions into the patent frameworks of South Asia brings a mix of challenges and prospects. By implementing well-considered legal reforms and fostering regional cooperation, South Asian nations can protect their innovation landscapes while positioning themselves at the forefront of the fast-changing field of AIrelated intellectual property. The evolution of patent law in this region will rely on a proactive approach, flexibility, and dedication to promoting inclusive and sustainable advancement in technology.
BIBLIOGRAPHY
- World Intellectual Property Organization (WIPO), WIPO Technology Trends 2019:
Artificial Intelligence, available at: https://www.wipo.int/publications/en/details.jsp?id=4386
- Thaler v. Hirshfeld, 558 F. Supp. 3d 238 (E.D. Va. 2021), aff’d, Thaler v. Vidal, 43 F.4th 1207 (Fed. Cir. 2022).
- UK Intellectual Property Office, Decision on the Patent Applications in the Name of Stephen L. Thaler, BL O/741/20 (21 Sept. 2020), available at: https://www.ipo.gov.uk/pchallenge-decision-results/o74120.pdf
- Commissioner of Patents v Thaler [2021] FCA 879 (Federal Court of Australia).
- South African Companies and Intellectual Property Commission, Patent Journal No. 5, Part 1 (July 2021).
- European Patent Office, Legal and Practical Aspects of AI-Inventorship: DABUS Applications, available at: https://www.epo.org
- Rajiv Aggarwal, Artificial Intelligence and Intellectual Property in India: Challenges and the Way Forward, Centre for Internet & Society, 2022.
- Bhavani Prasad Panda, Patentability of AI-Generated Inventions in India: A Legal Vacuum?, NALSAR Law Review, Vol. 17 (2022).
- Shirin Banerjee, Redefining Inventorship in the Age of AI: South Asia’s Legislative Lag, Journal of Intellectual Property Rights, Vol. 27 (2022).
- SAARC Intellectual Property Rights Cooperation Framework, SAARC Secretariat, Kathmandu, 2020.
- Saurabh Ghosh, Policy Considerations in Patent Law Reform for AI-Driven Innovation, Indian Journal of Law and Technology, Vol. 18 (2023).
Name- Kashika Dubey
College – Amity University Madhya Pradesh, Gwalior
Topic- PATENTABILITY OF AI-GENERATED INVENTIONS:
CHALLENGES IN HARMONIZING SOUTH ASIAN IP LAWS
Enrollment No- A61021520048
[1] Abbott, R., “I Think, Therefore I Invent: Creative Computers and the Future of Patent Law,” Boston College Law Review, 57 (2016): 1079–1126.
[2] Rimmer, M., “AI as Inventor? DABUS, Patents and the Shape of Things to Come,” European Intellectual Property Review 43, no. 5 (2021): 261–265.
[3] Ryan Abbott, The Reasonable Robot: Artificial Intelligence and the Law (Cambridge University Press, 2020), pp. 112–118.
[4] World Intellectual Property Organization (WIPO), “DABUS: Artificial Intelligence System as Inventor,” WIPO Magazine, December 2021, available at:
https://www.wipo.int/wipo_magazine/en/2021/06/article_0004.html.
[5] Australian Federal Court, Commissioner of Patents v Thaler [2021] FCA 879; Thaler v Commissioner of Patents [2021] FCA 1302; see also Ryan Abbott, “The DABUS AI and Inventorship,” Journal of Intellectual Property Law & Practice 16, no. 10 (2021): 885–889.
[6] WIPO, Intellectual Property and Artificial Intelligence: A WIPO Perspective (Geneva: World Intellectual Property Organization, 2020), 45–48
[7] R. K. Singh, “Patent Law Challenges in South Asia: Addressing AI-Generated Innovations,” Journal of Intellectual Property Rights 26, no. 3 (2021): 134–142
[8] N. Hussain and A. Rahman, “Harmonization of Patent Laws in South Asia: The AI Imperative,” South Asian Journal of Legal Studies 9, no. 1 (2023): 67–75.
[9] Abbott, Ryan. The Reasonable Robot: Artificial Intelligence and the Law. Cambridge University Press, 2020.
[10] Bently, Lionel, and Brad Sherman. Intellectual Property Law, 5th Edition. Oxford University Press, 2020