Domestic Violence - A Social Evil

Author: LawSphereX

Category: General Law

Published: 2026-08-23T19:34:04.712Z

Updated: 2026-08-30T07:09:22.940Z

AI AS AN INVENTOR? NAVIGATING THE LEGAL

CHALLENGES

OF AI AND PATENT LAW

Priya Elizabeth Sunil, BA LLB (Hons), NUALS

ABSTRACT

The emergence of AI in the current gen-ze era has had several implications for the youth and the working class, who have taken this as a boon to help carry out their daily tasks. This system has not only lent a helping hand in automating tasks that were once explicitly performed by humans but has also made significant strides in content creation where machine algorithms are used, ranging from content creation and music to generating automated articles. This has deranged people from using their creative hats and made them lethargic when performing tasks effectively. This paper aims to scrutinise such questions as whether AI can be made an inventor where it was once exclusively placed in the hands of humans.

INTRODUCTION

In the rapidly evolving landscape of the 21st century, seeds of innovative technology have been planted throughout the years, leading to the evolution of technology per se, bringing artificial intelligence into light. 1This change has become an integral part of our daily lives, revolutionising industries and challenging the traditional notions of creativity and efficacy. 2 This has emerged as a game changer, intensifying the capability of increasingly sophisticated growth, which can generate a vast array of content, from music and artwork to many more. This unpredictable tune to technology and creativity has led to a new era of innovation, which has sparked a debate in intellectual property (IP) law and otherwise. This very system has captivated the minds of humans in several ways, embarking on the suppression of innovative methods that lead to critical thinking and creativity.

This technology has outpaced the industry, shaped our daily lives and captured replicated human knowledge. The phrase “Blurring the lines” between human and artificial creativity symbolises the enormous volume of machine-created works that increases the gaps between the former and the latter. This has posed questions on the fundamental principles of copyright law, patent protection and trademark rights- systems traditionally designed to foster and protect human ingenuity. As we stand at the cusp of this technological revolution, it becomes imperative to determine the legal implications to concur with the challenges that arise and help resort to a more defining solution. This paper aims to delve into the depths of AI law and the intricate relationship between AI and intellectual property rights, evaluating the tensions within the same.

This paper also aims to delve into the intricacies of demystifying AI by tracing its evolution from the early days of its invention and understanding the underlying principles that have propelled AI’s development to what it is today. One of the most pertinent issues raised in light of intellectual property rights is “The patentability of AI-generated Inventions,” which also enunciates the potential for AI systems to be recognised as inventors or co-inventors. Thus, we shall be shedding light on the 1 IIPRD, Intellectual Property Rights in the Age of Artificial Intelligence: Navigating Challenges and Seizing Opportunities,https://www. iiprd. com/intellectual-property-rights-in-the-age-of-artificial-intelligence-navigatingchallenges-and-seizing-opportunities/ (last visited June 18, 2024).

2 Vishnu S, Navigating the Grey Area

Copyright Implications of AI Generated Content, 29 Journal of Intellectual Property Rights 103-108 (2024) implications surrounding AI patents and patents by AI by addressing these issues within IPR. That said, it shall also explore the potential solutions to address the legal challenges incorporated by AI and the promotion of transparency and accountability in AI development. By examining these issues, we hope to contribute to the crucial debates surrounding the future of intellectual property rights in which the age of the boundaries between human and machine creativity is increasingly blurred. EVOLUTION OF AI: A HISTORICAL BACKDROP In a world where machines have taken control over the majority of the tasks that make the lives of humans easier through rapid advancement in the realm of artificial intelligence, technology has outnumbered.

The “AI systems” that generate innovative solutions to problems have skyrocketed the worries about the existing patent law, designed primarily to place human inventions as the crux of the framework. The pertaining system faces difficulties in determining inventorship and ownership. AI has evolved meticulously to give birth to Large Language Models (LLMs), a milestone in the evolution of AI, embarking on its ability to understand and generate natural language with a similar level of sophistication that was once thought impossible. Such a technology mimics human intelligence to its maximum. It can also engage in coherent conversations, generate creative content and assist in complicated tasks, blurring the lines between human and machine capabilities.

The major question that has popped up is whether the system designed to foster human ingenuity is flexible enough to incorporate “non-human inventors? ” And should it be considered in the first place? Copyright law has not existed beyond the human touch; it aims to protect the original works of Artists i,e, authorship, granting them the exclusive rights to use their content to the fullest and distribute their creations. However, as soon as machines have stepped in as content creators, this has attracted complications. The use of AI has not taken heat for long as the relevance of AI in IP law has a relatively short history. Even then, AI has been used consistently since the 2000s and persistently helping rapidly in fields like healthcare and finance. Since then, AI has been evolving tremendously, which has led to its increased use in Intellectual property.

One of the biggest hits in AI usage was in the field of patent law. Just like how AI has helped the healthcare and finance sector, it has also been useful in analysing patent databases to identify the relevant patents for new inventions. Since then, AI has acted as an intermediary between human inventions and creativity and has also expanded the use of AI in other areas, such as copyright and trademark law. That said, the rise in the use of computing power is “poised to take creative machines from novelties to major drivers of economic growth. 3 This phenomenon poses new challenges, loosening the strengths of adaptability between humans as inventors and making computers more apt in generating patentable subject matter under circumstances in which the computer outperforms humans and qualifies to meet the requirements of being an inventor.

4 Moreover, legally speaking, there is no adequate case law in proving the qualification of a computer as an inventor of a patentable subject matter, and thus, no statute nor patent office policy5 addressing the same. What is the role played by AI? AI-generated content is exclusively created using machine algorithms that generate data using 3 5 Of perhaps greater concern than automation displacing human workers, a number of prominent scientists are concerned about the implications of machines outperforming people. Professor Stephen Hawking has warned that self-improving computers could threaten the very existence of humanity. Rory Cellan-Jones, Stephen Hawking Warns Artificial Intelligence Could End Mankind, BBC (Dec. 2, 2014), http://www. bbc. com/news/technology-30290540 [https://perma. cc/2NWE-D8MB]. Chris Isidore, Elon Musk Gives 0M to Fight Killer Robots, CNN Money (Jan.

16, 2015), http://money. cnn. com/2015/01/15/technology/musk-artificial-intelligence/ [https://perma. cc/6N3U-WTWF]. Dylan Love, Scientists Are Afraid to Talk About the Robot Apocalypse, and That’s a Problem, Business Insider (July 18, 2014), http://www. businessinsider. com/robot-apocalypse-2014-7#ixzz3QQpLJ0Jj [https://perma. cc/AH5J-5G34]. Kevin Rawlinson, Microsoft’s Bill Gates Insists AI Is a Threat, BBC (Jan. 29, 2015), http://www. bbc. com/news/31047780 [https://perma. cc/9MXD-TQ68]. Irving J. Good, Speculations Concerning the First Ultraintelligent Machine, 6 Advances in Computers 31 (1966). 4 Stephen L. Thaler, Synaptic Perturbation and Consciousness, 6 Int'l J. Machine Consciousness 75 (2014) [hereinafter Thaler, Synaptic Perturbation and Consciousness].

Stephen Thaler, Creativity Machine® Paradigm, in Encyclopedia of Creativity, Invention, Innovation, and Entrepreneurship 451 (Elias G. Carayannis ed. , 2013) [hereinafter Thaler, Creativity Machine® Paradigm]. S. L. Thaler, A Proposed Symbolism for Network- Implemented Discovery Processes, World Congress on Neural Networks ’96, San Diego 1265 (Int'l Neural Network Soc'y 1996) [hereinafter Thaler, A Proposed Symbolism]. 5 Ben Hattenbach & Joshua Glucoft, Patents in an Era of Infinite Monkeys and Artificial Intelligence, 19 Stan. Tech. L. Rev. 32, 44 & n. 70 (2015). Robert Plotkin, The Genie in the Machine 60 (2009). About Us, USPTO, http://www. uspto. gov/about-us [https://perma. cc/6HZY-V9NU] (last visited Jan. 27, 2016). the same technology. 6 They are designed so that content is generated in the fastest way possible compared to humans.

Machine-based learning can be classified into two: - Supervised learning - Unsupervised learning The former is trained on labelled datasets where the input data is already classified. 7This system helps in the pre-classification of data with the algorithm i,e by using the labelled data to learn how to generate new content similar to the previous examples in the dataset. A threestep process follows the same: training, learning, and data generation. Firstly, the algorithm is fed a dataset where each input is paired with its correct output. Secondly, by analysing these input-output pairs, the algorithm recognises the patterns and the relation between the inputs and their labels. Lastly, once trained, the algorithm can now generate new content that follows the earlier patterns learnt, and based on the specific inputs, it creates new content.

For instance, a learning algorithm trained on a large dataset of news articles, each labelled with its topic, can learn to generate new articles on specific subjects with the requisite keywords. Moreover, the latter works with unlabelled datasets without any pre-defined categories, and such algorithms use methods like clustering and dimensionality reduction methods within the data independently. 8A three-step process follows: Pattern recognition, structure inference and Content Generation. Firstly, using the aforementioned methods, the algorithm keenly analyses the dataset to identify the inherent patterns. Secondly, based on the patterns discovered, the algorithm infers a structure within the data. Lastly, the algorithm can generate new content that adheres to the discovered patterns using the inferred structure.

That said, reflecting on the pertinent question of ownership and authorship of AI-generated content raises an impediment. The role of machine learning algorithms in generating AI content poses a procedure9 for processing vast quantities of data and generating new content based on 6 Lund B. D. , Wang T. , Mannuru N. R. , Nie B. , Shimray S. , & Wang Z. , ChatGPT and a New Academic Reality: Artificial Intelligence-Written Research Papers and the Ethics of the Large Language Models in Scholarly Publishing, 74 Journal of the Association for Information Science & Technology 570 (2023). 7 Kate Crawford & Trevor Paglen, Excavating AI: The Politics of Images in Machine Learning Training Sets, 36 AI & Society 1105 (2021).

8 Majed Binkhonain & Liang Zhao, A Review of Machine Learning Algorithms for Identification and

Classification of Non-Functional Requirements, 1 Expert Systems with Applications: X 100001 (2019). 9 Philip Hacke, Andreas Engel & Michael Mauer, Regulating ChatGPT and Other Large Generative AI Models, arXiv preprint arXiv:2302. 02337 (2023). previous data. Moreover, these algorithms use complicated mathematical models to recognise the patterns10, and the algorithms themselves cannot do this; they need assistance as algorithms do not have a creative natural intent, and they can only create a similar variation of it like human-developed works. Thus raising questions regarding the ownership and copyright law. 11 THE RISE OF AI IN COPYRIGHT LAW Copyright law protects the artists or the authors of their original work to use and distribute.

This does not work well with AI-generated content and traditional copyright law due to the complexities that arise on who owns rights to the content created by machines. 12 Most jurisdictions do not accept AI-generated content by machines as authors, and the debate is still in an uproar. 13 one significant way to determine the faults in AI-generated content is Novelty. In the case of Gaurav Bhatia v Union of India, the Delhi high court held that AI-generated inventions could only be patented if they met the novelty category. Unlike machines, the human brain has room for immense variants of creativity through novel ideas using innovative ways that require skill in art. However, machines, on the other hand, can process massive amounts of data.

They can even generate solutions that may be new, but determining whether AI is purely novel is a hard task as it uses transit solutions from the previous data and does not use novelty. Thus, the question of AI is more complex as AI creates content designed to make judgements based on patterns lacking creative touch. Beyond the role of non-obviousness and novelty, there lies the active involvement of human input in the content-creating process14 i,e. the AI systems can generate the outputs, the option of whether they must be considered as the sole investors in the project is a question that takes an example where:- if a human provides assistance or guidance or any form of supervision to the machine algorithm during the content creation process then they may be considered as coinvestors or co-authors to work.

This has been a recurring debate along the lines of patent 10 Xin Shu & Yanjun Ye, Knowledge Discovery: Methods from Data Mining and Machine Learning, 110 Social Science Research 102817 (2023) 11 Jurate Kalpokiene & Ignas Kalpokas, Creative Encounters of a Posthuman Kind – Anthropocentric Law, Artificial Intelligence and Art, 72 Technology in Society 102197 (2023). 12 Debdip Jodha & Pritha Bera, Copyright Issues in the Era of AI: A Critical Analysis, 13 Res Militaris 1737 (2023

13 Marcin Miernicki & Iva Ng, Artificial Intelligence and Moral Rights, 36 AI & Society 319 (2021).

14 Nadiah Selvadurai & Rūta Matulionytė, Reconsidering Creativity

Copyright Protection for Works Generated Using Artificial Intelligence, 15 Journal of Intellectual Property Law & Practice 536 (2020). offices and courts grappling with the same regarding inventorship by AI and its significance. This issue has not been on standby, and a few countries have gotten leads on addressing it. For instance, the European Union’s copyright Directive includes provisions that clarify the legal status of AI-generated content and ensure they are adequately compensated.

LEGAL

CHALLENGES

POSED BY AI-GENERATED INVENTIONS One of the overriding aspects that must be considered is the license and ownership of the content created by AI. The algorithms used in creating the content may belong to humans or organisations holding copyright. This poses questions on whether AI solely owns the copyright for the same. That said, AI-generated systems hold human support as the system can be owned and operated by an individual or a company per se; this can make the former or the latter the copyright owner. However, there are instances where the AI system is trained effectively on the copyrighted material, which makes it difficult to determine the ownership, thus making it complex. 15 This defines the phrase “ blurring the lines between human and machine capabilities”, where it makes it difficult to determine who exerted the creativity and human judgement in the contents.

Therefore, it is important to establish clear guidelines regarding the ‘Copyright of AIgenerated content’ to address the hindrances associated. Encouraging the same, it must ensure that the content developed by AI is equipped with all the necessary licenses and permissions which have to be obtained. Since there has been a swamp of concerns regarding originality, which is necessary to qualify for copyright protection, unlike AI-generated content, the copyright law in most jurisdictions has mandated that a human author must create the work, thereby precluding machine-generated content. 16 A pertinent example citing the same is in the United States, where the Copyright offices have explicitly stated that they shall not register works created by non-human inventors, thus denying attributing inventorship to machinegenerated works.

There are ongoing debates in India on the same, yet lacks clarity as there is no clear precedent or policy from the Indian Copyright Office. In 2020, there was an incident where the application on AI (RAGHAV), being the sole investor, was rejected. Still, another 15 Yogesh K. Dwivedi et al. , “So, What if ChatGPT Wrote It? ” Multidisciplinary Perspectives on Opportunities, Challenges, and Implications of Generative Conversational AI for Research, Practice, and Policy, 71 International Journal of Information Management 102642 (2023)

16 Clark D. Asay, Independent Creation in a World of AI, 14 FIU Law Review 201 (2020).

application with humans and AI as co-authors was granted permission without a clear explanation. The Copyright Act of 1957 requires a certain level of creativity or novelty in the content creation. However, there is no clear definition on the definition of “person” in the copyright act, making it difficult to comprehend the same. This ignites challenges in determining ownership and copyright protection for AI-generated content. Moreover, AIgenerated content draws content from several sources that may be copyrighted, thereby infringing on the rights of other creators. If you take an example, YouTube content creators and other article makers may include songs or even quotes from other content creators, thus raising questions on the aspect of “fair use”. The word ‘inventor’ has certain legal implications that do not affect the invention's patentability.

Patent law is different because the patentability of an invention is assessed based on the objectivity test and not subjectively. Take a vivid example from the High Court of Australia, which has observed that a valid patent “may be obtained from something stumbled across by accident or remembered from a dream. ” Such examples from various jurisdictions have viewed AI-generated content in different ways. In traditional intellectual property law, the author is the sole creator of the work and is entitled to the rights and protections for the same, but in the case of AI-generated, it’s unclear. Moreover, an inventor has traditionally been defined as a natural person who conceives of an invention.

However, the Indian Patents Act, 1970, section 6(1)(a) does not explicitly define the term ‘inventor’ but however talks about the eligibility for a person to apply for a patent, where he or she must be the “true and the first inventor” of the invention. Similarly, in the United States, the Patent Act refers to inventors as “individuals”, which has been interpreted to mean natural persons. Additionally, section 2(1)(y) specifies that “true and first inventor” does not include either the first importer of an invention into India or a person to whom an invention is first communicated from outside India. ” While in section 2(1)(y) it encompasses the individuals or entities who are ineligible to be recognised as investors, where it opens room for ambiguity when determining the true and first inventor of an invention where AI is the original owner of the content.

Such a situation ignites cumbersome attributes in identifying the true owner of the content created. In a pertinent case of Som Prakash Rekhi vs Union of India & Anr17, the supreme court held that a “jurisdictional person is the one to whom the Law

17 Som Prakash Rekhi v. Union of India & Anr, 1981 A.I.R. 212, 1981 S.C.R. (2) 111 (India).

attributes ‘personality’. When we refer to a juristic personality, we refer to a legal entity that has the right to sue, or that can be sued by another entity”. Since AI is not a legal entity with the right to sue, the court held that AI could not independently exercise its rights or fulfil any legal entity's responsibilities. This discrepancy raises crucial questions about whether patent law can accommodate AI as inventors without undermining the crux of the principles of patent law. The global aspect that sought to achieve a solution in defining the legality of whether the AI system is capable of being named as the inventor of patentability is subjected to various tests across the world, and such a test is the DABUS test, which was a project by Dr Stephen Thaler.

GLOBAL APPROACHES TO AI INVENTORSHIP Determining the capability of ‘AI patents and patents by AI’ is a subject matter that is tested globally across different jurisdictions as part of ‘The Artificial Inventor Project. ” This project is accompanied by various tests which help foster the same. DABUS, also known as Device for the autonomous bootstrapping of unified Sentience, is a system classified as ‘connectionist AI. ’ Its architecture consists of two main components:-

1. A network of neural networks that generates novel ideas

2. A separate neural network system that evaluates these ideas for originality.

Through the aforementioned two-step process, DABUS has independently created two inventions

1. A fractal container designed for food storage

2. A neural flame intended for use as a search and rescue beacon

In this case of Thaler v Vidal18 it was held that the DABUS test did not qualify as an inventor because it failed to fulfil the criteria of being classified as a natural person or a human under US law. Therefore, US law held that there must be both intentionality and culpability, which

18 Thaler V Vidal, No. 21-2347 (Fed. Cir. 2022).

AI lacks as it cannot enter into legally binding contracts and thus cannot categorise AI as a sole inventor. The global trend in courts with varying opinions from country to country has rejected applications stating that AI cannot be an inventor for the purpose of patent law. This has not only been the case in the US court of law but other opinions in the Australian court, United Kingdom, where both the high court and court of appeal agreed that the UK comptroller of patents was correct in rejecting the DABUS test. It had to be a natural person. Moreover, in a separate case involving the DABUS application, the 18th Senate of the Federal Patent Court reached a different conclusion. They ruled that patents for AI-generated inventions could be granted, but only if the applicant removes any mention of AI in the inventor designation.

Currently, only two countries, South Africa and Saudi Arabia, stand as the only jurisdictions diverging from the global consensus. Dr. Thaler has also expanded his efforts by submitting applications to other countries, including China, Japan, Switzerland, and more. It remains uncertain whether the tables will turn, which would further form exceptions to the global trend. A NOVEL TAKE My perspective on AI as an inventor shifts the focus from the binary question of whether AI can be an inventor to a more nuanced approach to how patent law can evolve to accommodate or incorporate the new realities of human–AI collaborative innovation. We should reframe the debate and focus more on the collaborative aspect instead of asking, “Can Ai be an inventor? ” we must shift its focus to “ How can patent law evolve to reflect a modern collaborative innovation process? ”.

This shift allows us to address the issues more effectively. The light to inventorship is PSA (a person skilled in the art), which AI cannot accompany due to its lack of creativity; thus, incorporating both human art and AI can create a detailed approach, making both AI and human co-authors. In the case of South Asia FM Limited v Union of India, the Delhi high court held that the song, which was considered a work under the Copyright Act of 1957, the court held that the song lacked human creativity and was not a result of novelty. Therefore, for a work to be eligible for copyright protection, it must result from human creativity and an essence of originality. It is also said that a natural person creating an invention must contribute to it effectively, which was specified in the case of Pannu v. lolab Corp 19.

The natural person must have significantly contributed to each claim in the patent application. In contrast, a single person who uses an AI system to create an invention must contribute to every claim in the application to the fullest. In such a way, it draws down the complexities arising in debates about AI as inventors or whether they must be incorporated.

CONCLUSION

The global chaos surrounding the determination of AI inventorship in patent law poses several questions across jurisdictions. It underscores the pressing need for a more flexible shift in our approach to intellectual property rights. The divergent rulings and the exceptions poised in the DABUS cases signal the importance of the recurring debate in the face of rapidly advancing technologies. As we navigate through the branches of the complexities, it is crucial to strike a definite balance between fostering innovation, creativity, and novelty and protecting individual rights while serving the broader interests of society. To effectively carry out the co-authorship between AI and human creativity, it is necessary to balance both by not letting AI outperform human touch.

The ongoing debates across various jurisdictions offer a unique opportunity to tune in with different opinions and embrace the potential of AI, which preserves the fundamental aspects of patent law. Therefore, by adopting a more ethical and practical approach to patent law, we can not only create a well-equipped system but also accommodate the current challenges posed by AI and reflect in the face of future technological advancements.

19 Pannu v. Iolab Corp., 155 F.3d 1344, 1351 (Fed. Cir. 1998).