Is Joint Authorship Necessary Under The Current Copyright Law For Human-Assisted AI Creations

Author: LawSphereX

Category: Technology, Cyber & Privacy

Published: 2026-08-23T20:24:03.906603+00:00

Updated: 2026-08-26T17:10:58.901+00:00

IS JOINT AUTHORSHIP NECESSARY UNDER THE

CURRENT COPYRIGHT LAW FOR HUMAN-ASSISTED AI

CREATIONS?

Yash Vikram Singh, LLM, National Law School of India University

ABSTRACT

This paper deals with the legality of providing joint authorship to the duo of a human being and an Artificial Intelligence (AI). The paper will limit itself to current legal regulations, finding cues for validating AI authorship from copyright legislation of countries that have dealt with the issue and narrowing down the discussion to academic and scientific research fields and the role AI can play in a collaborative sense. The paper will discuss the ambiguity in copyright laws, wherein some of the territorial laws stay silent on the need for an author to be a human being. The paper will conclude with the scope of co-authorship and the benefits to society, if any.

Introduction With every passing decade, the rate of technological innovation has accelerated exponentially, and new methods of content creation have emerged, particularly with the rise of AI-generated content, raising questions about copyright ownership. Can AI be considered the owner of its creations, or does this right belong to the person who developed the AI? In general, as per the majority of the copyright laws, the creator of a work holds the copyright. This brings forth the question: can AI be regarded as the copyright owner of its generated content? This paper will make an effort to answer this question in a concise and legally sound manner. As Agnes1 mentions the argument of the two sides, i. e. , AI is the author and owner of its creations versus AI lacks authorship and merely functions as a tool programmed by humans.

The very fact that the generative AI relies heavily on data input provided by its developers or prompter, its authorship and ownership in content creation, continues to raise complex legal and ethical challenges, requiring further exploration. The foundational inquiry to be discussed in this paper is not novel per se, as authors like Wenqing2 already discussed terms such as “Machine Authorship”, wherein he argues how human authorship can be equated with the AI’s ability to generate copyrightable content. It reflects human creativity expressed through AI, which functions as a tool that carries out its creator's instructions. Any originality or creativity produced by the algorithm should be attributed to the developers as co-authors. Thus, AI authorship can be viewed as an extension of human authorship, exercised indirectly through the machine.

The concept of joint authorship can be viewed from two opposing perspectives: one argues for granting AI authorship to distinguish between human-generated content and AI-assisted work, while the other asserts that AI is merely a manifestation of human intellect. The creativity of AI ultimately belongs to humans3, who dictate and control the ideas of AI, albeit with the aid of hardware that exponentially enhances the speed of its idea processing. The algorithm behind 1 Agnes Augustian, Authorship of AI-Generated Works: An Analytical Study, 4 Indian J. L. & Legal Res. 1, 1–13 (2022–2023). 2 Wenqing Zhao, AI Art, Machine Authorship, and Copyright Laws, 12 Am. U. Intell. Prop. Brief 1, 1–18 (2020). 3 J. M. Balkin, Lecture: 2016 Sidley Austin Distinguished Lecture on Big Data Law and Policy: The Three Laws of Robotics in the Age of Big Data, 78 Ohio St. L. J. 1217, 1217–1241 (2017).

AI-generated art originates from human behaviour, learning and imitating artworks. This process mirrors what a traditional artist might do when starting to create art, but AI performs the same task at a much faster rate. None of the above-mentioned arguments would hold water if the intention of the people involved with respect to authorship is not known4. The editor of any publication also assists the writer in the final draft, so how far can the analogy be stretched to joint authorship? Herein, the intention of both parties from the very beginning of the project becomes quaint-essential. The following has been dealt with in detail with respect to the Indian Copyright Act5 by the Hon’ble Delhi HC in Najma Heptulla vs. Orient Longman6.

Niching down the above-mentioned discussion to the academic, scientific research, and visual arts fields would provide us with a framework to improvise further in other fields for two primary reasons. One, both fields form the crux of any given field, be it business, entertainment, or personal research. Two, there is already a head start in the scientific field with respect to labelling AI-generated content, and the academic field stirred the pot with strong friction for AI-generated academic content, majorly contesting the originality and plagiarism aspects of the content generated. One research paper may not answer all the questions related to the intricacies of co-authorship, but setting the foot in the right direction is the need of the hour.

However one thing is clear, the current legal paradigm cannot inculcate AI without a nexus with humanness in order to align the idea of co-authorship within the prevailing copyright legislation. The Human-AI authorship paradigm All that AI is creating now In the mid-20th century, when the Research and Development (RAND) Corporation funded the Logic Theorist7, a designated program for iterating human beings’ problem-solving methods 4 Laura Biron & Elena Cooper, Authorship, Aesthetics, and the Artworld: Reforming Copyright’s Joint Authorship Doctrine, 35 Law & Phil. 55, 55–86 (2016).

5 Copyright Rules, 1957

6 AIR 1989 Delhi 63, 1988 (2).

7 Rich Anyoha, The History of Artificial Intelligence, Science in the News—Harvard University (Aug. 28, 2017), https://sitn.hms.harvard.edu/flash/2017/history-artificial-intelligence/.

via machine, gave birth to artificial intelligence, and today the scientific field stands divided in its rightful sense on the question of copyrightability of AI-generated content. The world has come a long way from manifesting a program like AI to present times, where we are discussing the copyright protection for AI, wherein AI is composing music with cues from past renditions or half-composed works.8 People are able to make statements like, “I’m making thousands using AI to write books”, Tim Boucher9 has published over 97 books since

2022 with the aid of AI. Some high-profile journals like Elsevier’s Surfaces and Interactions10

have already published AI-assisted papers; the journal further clarified its position by mentioning that they promote self-declaration of AI assistance from the authors’ side. AI is assisting in the medical literature; Qure. ai11 is assisting in machine learning-based conclusions of CT scans, which align with the conclusions of radiologists in 89% of cases. Countries like Nigeria are setting foot in the right direction, having established the National Centre for Artificial Intelligence and Robotics (NCAIR)12, a special-purpose vehicle created for the development of emerging technologies in Nigeria. Current legal battles All the positive nods for AI copyright protection or at least the recognitions’ above-mentioned are not sailing through the wind without resistance. There is a whole conundrum of legal battles ensued by different parties.

The core of the disputes with respect to AI and copyright infringement arise primarily due to the humongous datasets used for training the modal. The protesting parties believe that all this usage should be done post-licensing or compensation, as would be feasible13. 2022 saw a class-action suit14 against GitHub and Microsoft OpenAI by a group of programmers, primarily on grounds of privacy concerns and breach of contract.

8 Adam Brooksbank, Music Made by Machines? Inside the Latest Revolution in Music, NBC News (May 29,

2019) [Video],

9 Thomas Boucher, “I’m Making Thousands Using AI to Write Books”, Newsweek (May 15, 2023),

https://www.newsweek.com/ai-books-art-money-artificial-intelligence-1799923. 10 Michael Coddington, Scientific Journal Publishes Paper with AI-Generated Introduction, Technology Networks (Mar. 21, 2024).

11 About Us, Qure.ai, https://www.qure.ai/about-us.

12 National Centre for Artificial Intelligence and Robotics, https://ncair.nitda.gov.ng/about-us.

13 A.J.L. Roy, Legal Battles Over AI and Copyright: Emerging Trends in Intellectual Property Disputes, Passle (June 5, 2024). Garry Lawton, Is AI-Generated Content Copyrighted?, TechTarget (June 5, 2024), https://www.techtarget.com/searchcontentmanagement/answer/Is-AI-generated-content-copyrighted.

Getty Images15 goes on to the London High Court to sue Stable Diffusion AI for copyright infringement of more than 12 million copyrighted images from its platform. Another lawsuit in San Francisco saw author Sarah Silverman16, accusing OpenAI of copyright infringement for making use of the author's book, Bedwetter, claiming that without the use of content from the book, OpenAI could not have trained itself for the given subject. Aligning the legal framework for copyright privileges for AI Authorship Generally, an author is a person who has invested their creative and intellectual effort in the production of the copyrighted work. This does not include merely stenography or mechanical effort, as mentioned in the Donoghue17 or the Leah18 cases, where authorship was defined more strictly to emphasize originality and creativity.

In India, Section 2(d) of the Copyright Act mentions the word person in two of the subsections, first and last; hence, a “person” is essential in the completion of any of the works mentioned in the section. Similarly, in Nigeria, Section 51 of the Copyright Act19 iterates the same clauses as that of India; interestingly, they mention “person” in every single sub-clause. In the United States, Section 201 of Title 17 of the US Code20 talks about authorship of copyrights with respect to writers, and work for hire, and 201(b) mentions that the employer may become the owner but the writer is the author. 21 In the UK, Section 9 of the Copyright, Design, and Patents Act, 1988,22 is similar to the Indian Copyright Act, with the same bifurcations and labels, mentioning the person twice in 15 James Vincent, Getty Images Sues AI Art Generator Stable Diffusion in the U. S.

for Copyright Infringement, The Verge (Feb. 6, 2023)

16 Donoghue v. Allied Newspapers Ltd., [1938] 1 K.B. 578.

17 Ibid

18 Leah v. Two Worlds Publ'g Ltd., [1951] Ch. 393.

19 Nigerian Copyright Commission, Copyright Act, https://lawsofnigeria.placng.org/laws/copyright.pdf (1988).

20 U.S. Copyright Office, Copyright Law of the United States, https://www.copyright.gov/title17/.

21 17 U.S.C. § 201.

the section. The term author may not be clearly mentioned as “human” in some of the sections, although the copyright Acts mentioned above always emphasize that copyright can be granted in works originated by humans alone. The propositions such as “work for hire” can be used to include AI within the ambit of authorship, as that will abide the human for liability and accreditation to AI for the work it will produce. Although the problem with the “work for hire” analogy with AI is that such works should have written agreements, as is the mandate23, so it will lead to further legal proceedings. Concept of joint authorship in copyright law There is a plethora of literature on the subject of joint authorship; as per the de minimis test prescribed in the Feist Publication24 case law, AI could meet the benchmark.

The independent creation by AI based on prompts or past collected data would qualify as creative creation. This will surely need some tweaks to our current copyright systems, but will ensure that we protect both the interests of the public and accreditation to the AI assisted systems in use25. Unsuccessful attempts at proving AI authorship There can be no paper that can proceed explaining non-human authorship without explaining the Monkey-Selfie (Naruto vs. Slater)26 case law, rejecting the copyright grant claim stating that the monkey may have a constitutional backing but the statutory claim finds no backing as per the Hon’ble Judges of the US Supreme Court. Hon’ble Justice Smith statement27, “The court can only construe legislation and cannot itself legislate, no matter how great the policy needs.

” in the DABUS-Stephen Thaler case, provides an in-depth understanding that, as long as humans are not involved in the copyright grant, there can be no sole rights to any non-human entity over any piece of copyright. 23 Crown, LLP, What Is a Work for Hire? , Crown LLP (Dec. 2, 2021), https://crownllp. com/blog/what-is-a-workforhire/#:~:text=Second%2C%20the%20work%20must%20have,%E2%80%9Cwork%20made%20for%20hire. %E 2%80%9D.

24 Feist Publ’ns, Inc. v. Rural Tel. Serv. Co., 111 S. Ct. 1281 (1991).

25 Nathan Shoemaker, Japanese AI Writes a Novel, Nearly Wins Literary Award, Big Think (Sept. 30, 2021), https://bigthink.com/technology-innovation/a-japanese-ai-wrote-a-novel-almost-wins-literary-award/.

26 Naruto v. Slater, 16-15469, (9th Cir. Apr. 23, 2018).

27 Théâtre D'opéra Spatial, [2020] EWHC 2412 (Pat).

Scope of AI authorship in academic field One of the articles in a European journal28 provides two opposite scenarios, one wherein a paper got desk rejection due to over 50% usage of ChatGPT, while on the other, several papers made diligent use of AI and gave it due credit. Although the conclusion herein also relies heavily on the fact that AI is not creating something innovative, nor can AI be held liable for whatever it has produced or advised. There have been instances where AI has written fiction, such as a narration of a road trip inspired by Jack Kerouac29. In Community for Creative Violence vs. Reid 30, an author is defined as somebody who; “Actually, creates the work, that is, the person who translates an idea into a fixed, tangible expression entitled to copyright protection.

” The term “person” remains the biggest contention in all the discussions around AI authorship; even Professor Nimmer31 skips explaining the term in the context of AI authorship. Providing some context to the interpretation of legal entities in India, the Hon’ble Delhi HC in Rupendra Kashyap v. Jiwan Publishing House Pvt. Ltd. 32 denied the Central Board of Education (CBSE), the righ to claim authorship over exam papers. The authorship claim can only be made by a natural person in the Indian Copyright Act; the position was enunciated in Tech Plus Media Private Ltd. v. Jyoti Janda33. AI is already being used to generate content processed through conditional logic, making the output sound more human-like. It helps scholars summarize, organize, and present data coherently while also analyzing and synthesizing specific details for collaborators34.

However, recent claims35 of AI authorship are both intriguing and complex, especially when infringement 28 Anıl Balat & İbrahim Bahşi, May Artificial Intelligence Be a Co-Author on an Academic Paper? , European J. Therapeutics (2023), https://eurjther. com/index. php/home/article/view/1688/1364. Bryan Merchant, Can an

Write a Novel?, The Atlantic (Oct. 1, 2018), https://www.theatlantic.com/technology/archive/2018/10/automated-on-the-road/571345/.

30 Cmty. for Creative Non-Violence v. Reid, 490 U.S. 730 (1989).

31 David Nimmer, Nimmer on Copyright, Vol. 1, 5-5 (1997).

32 1994 (28) DRJ 286

33 (2014) 60 PTC 121.

34 Bernard Marr, Artificial Intelligence Can Now Write Amazing Content—What Does That Mean for Humans?, Forbes (May 13, 2019), https://www.forbes.com/sites/bernardmarr/2019/03/29/artificial-intelligence-can-nowwrite-amazing-content-what-does-that-mean-for-humans.

35 Did a Human or a Computer Write This?, N.Y. Times (Mar. 7, 2015),

https://www.nytimes.com/interactive/2015/03/08/opinion/sunday/algorithm-human-quiz.html.

issues arise, as distinguishing between human-made and AI-assisted content is becoming increasingly challenging. New scientific research assistant Balat & Bahsi36 discuss AI's role as an author, basing their findings on a survey of the scientific community. Their paper advocates for raising the bar on collaborative authorship standards while safeguarding human intellectual contributions, one of our most valuable assets. While AI may assist with creativity, true scientific innovation stems from the human mind's ability to analyze problems, evaluate solutions, and pursue innovation where necessary. As Balel mentions37, AI can accelerate this process but cannot replace the practical, cognitive abilities of humans. Large language models (LLMs) can effectively annotate, paraphrase, and filter research, but they are tools, not substitutes for the expertise of scholars.

The primary achievement for any scientific researcher is the authorship status they are provided via copyright laws38. The joint first co-authorship as a concept provides all the authors with equal accreditation, de-escalating any debate upon the credibility and authority of the concerned author in that paper. Currently, around 30% of articles can be identified as AI generated, while the other 70% get amalgamated into the ocean of research39. Embracing AI's benefits is important, but relying solely on it for scientific innovation is not foolproof. The expertise gained by scientists, medical practitioners, and technical experts through years of experience, both documented and undocumented, cannot be replicated by AI alone. This is where human intervention is crucial.

Joint authorship allows us to harness the best insights from the human mind while AI excels in data collation and identifying research gaps, which in this way is an easy task. Artistic fervor of AI The other important facet of copyrightable content is that the originality requirement from a 36 Alpaslan Balat & Ibrahim Bahşi, We Asked ChatGPT about the Co-Authorship of Artificial Intelligence in Scientific Papers, 29 Eur. J. Therapeutics e16 (2023), https://doi. org/10. 58600/eurjther1719. 37 Yakup Balel, View of the Role of Artificial Intelligence in Academic Paper Writing and Its Potential as a Co- Author: Letter to the Editor, Eur. J.

Therapeutics (2023), https://eurjther.com/index.php/home/article/view/1691/1371.

38 Nathan Efron, Joint First Authorship, 107 Clin. & Experimental Optometry 243 (2024),

https://doi.org/10.1080/08164622.2022.2146486. 39 Robert Grillo, The Rising Tide of Artificial Intelligence in Scientific Journals: A Profound Shift in Research Landscape, Eur. J. Therapeutics (Aug. 1, 2023), https://philarchive.org/archive/GRITRT-7.

human is of the minimal threshold, wherein one does not have to question the creative spark existing with the author. The Rembrandt painting is an amazing example, as provided by Mammen & Richey40, wherein the author gives a specific prompt for the required result. The aid from all the programmers and artists fine-tunes the sketch, but can this be enough to attribute authorship to the back-end team? The other important case, Burrow41 mentions, is that when AI is producing content, it is doing all the creative interventions as per the pre-saved data and algorithm, which has been fine-tuned over the years. The programmer envisioned the result? Was the result due to the creative spark of the programmer? Do they need to prove this element?

The Indian Copyright Office42 expressed a debatable opinion in the SURYAST artwork case by Ankit Sahni, stating that providing a prompt is akin to guiding a human artist when commissioning art. This indirectly supports the significant role AI played in creating the artwork. Théâtre D'opéra Spatial by Jason M. Allen faced a second rejection at the review board of copyright43, bifurcating the role of human beings and the impact of AI on the final output. Although, in both these above-mentioned cases, had there been involvement of a human coauthor, the scenario would have been much different. The creativity of the AI in both cases has not been questioned; it is the authorship on technical grounds that comes into the foreplay. An opposite take against the US copyright system can be seen at the Beijing Internet Court in the case of Li vs.

Liu,44 wherein the court mentions prompting as an essential element of an AI generated image, wherein the plaintiff’s (author) personalized expression was essential to giving shape to the final product. Hence, the perception with regards to Doctrine of Sweat & 40 Crystal E. Mammen & Carrie Richey, AI and IP: Are Creativity and Inventorship Inherently Human Activities? ,

14 FIU L. Rev. 275 (2020).

41 Burrow-Giles Lithographic Co. v. Sarony, 111 U.S. 53, 54-60 (1884).

42 Second Request for Reconsideration for Refusal to Register SURYAST, U.S. Copyright Office (Apr. 23, 2023), https://www.copyright.gov/rulings-filings/review-board/docs/SURYAST.pdf. 43 Second Request for Reconsideration for Refusal to Register Théâtre D'opéra Spatial, U.S. Copyright Office (2023), https://www.copyright.gov/rulings-filings/review-board/docs/Theatre-Dopera-Spatial.pdf. 44 Colin W. Savage & Joshua Rosenfeld, Diverging International Approaches to the Copyrightability and Authorship of AI-Created Works, DWT (Jan. 30, 2024).

brow45 varies in different jurisdictions, presenting us with a grey region within the copyright laws. Conclusion The discussion ensued upon provides us with some good answers to the questions posed on the one hand, while, on the other hand, certain other discussions leave us sour-mouthed. The fruition of this paper lies in the fact that solo authorship to AI was assumed to be a hurdle which the world is not ready to surpass as of now, so was the justification as cited in various examples. There is the element of predictability in how the assisted-AI works; alas, there can be hardly any reliance/dependability on the output of the work expected46. In general47, what we see is that neither China (Li v. Liu), USA (Zarya of the Dawn), UK, EU, nor India (Suryast) have crossed the golden threshold for treating AI as an author for a work.

The partnership of AI with a human is a synergy like none other; the combination of AI and humans at a chess game could collectively defeat all the other AI-powered bots or humans, indicating that coexistence with AI is in the best interest of the humans; hence, segregating the organic human content from AI-assisted content is quaint-essential.

45 Doctrine of Sweat of the Brow, The Column of Curae (Sept. 6, 2020),

https://thecolumnofcurae.wordpress.com/2020/08/18/doctrine-of-sweat-of-the-brow. 46 Jessica Harker, Science Journals Set New Authorship Guidelines for AI-Generated Text, Nat’l Inst. of Envtl. Health Scis.(Mar. 2023), https://factor.niehs.nih.gov/2023/3/feature/2-artificial-intelligence-ethics. 47 Colin W. Savage & Joshua Rosenfeld, A Chinese Court Takes on Generative AI “Creations” in Li v. Liu, DWT (Jan. 30, 2024).