The Invisible Pen - Legal And Ethical Implications Of AI Ghost-Writing In Intellectual Property Law

Author: LawSphereX

Category: Technology, Cyber & Privacy

Published: 2026-08-23T21:23:42.667555+00:00

Updated: 2026-08-26T17:05:38.213+00:00

THE INVISIBLE PEN: LEGAL AND ETHICAL

IMPLICATIONS OF AI GHOST-WRITING IN

INTELLECTUAL PROPERTY LAW

Aadish Vasudevan, Symbiosis Law School, Hyderabad.1 Ananya Agarwal, Symbiosis Law School, Hyderabad.2

ABSTRACT

An empty page has the power to overwhelm and dishearten the writer. At the end, it all comes down to that one thought that helps you channelize your idea into the perfect piece of work. However, a blank page may also lead the author towards an abbreviated route and employ Artificial Intelligence as their assistant or rather ‘ghostwriter’. This might seem to expedite the process for the author; however, the reader seems to be indirectly defrauded through this, considering the author claims absolute originality when, in reality, an external unknown entity assisted unofficially. This study aims to understand the various aspects that revolve around the concept of ghostwriting with respect to Artificial Intelligence, specifically. The study shall commence by laying a foundation for the concept of ghost-writing and its gradual evolution.

Further, it shall delve into the copyright framework of different countries and provide a comparative analysis of India with other countries, therefore, highlighting Artificial Intelligence’s recognition with the concept of ghost-writing. Additionally, the study shall outline the ethical and policy concerns revolving around the practice of ghost-writing. Hence, the core objective of this research is to underscore the evolving role of Artificial Intelligence in reshaping the practice of ghost-writing. Keywords: Ghost-writing, author, copyright, authorship

1 Aadish Vasudevan, fourth year student at Symbiosis Law School, Hyderabad.

2 Ananya Agarwal, fourth year student at Symbiosis Law School, Hyderabad.

INTRODUCTION

Imagine it is your best friend’s birthday, and you are expected to craft a heartfelt birthday message for them. However, owing to time constraints and a shortage of creative ideas, you choose to seek assistance from ChatGPT to compose it. The question that arises here is whether that birthday wish you conveniently signed under your name should actually be considered a message from you or may be subsumed under the category of ‘ghost-writing’. Essentially, ghost-writing is the act of producing or creating something in alignment with the intentions of the creator, yet attributed to a third-party creator. What needs to be highlighted here is the thin line of difference between ghost-writing and plagiarism that exists due to the lack of the author’s knowledge and consent in the latter, and the contrary in the former.

3 The common link, however, here is that neither of them is proudly announced, hence demonstrating a hint of falsity. Moreover, one of the most significant dilemmas that has ensued with the incorporation of this concept is the ambiguity between ‘author’ and ‘authorship. ’ It needs to be understood that ‘authorship’ is usually affiliated with recognition of creativity, while ‘author’ is someone who envisions the idea initially. 4 While ghost-writing is a practice that has sustained across centuries, hidden behind the shadow of confidentiality agreements, its interaction with Intellectual Property Rights (hereinafter referred to as “IPR”) brings light to the ethical, legal, and practical concerns that come along with it. Moreover, the introduction of Artificial Intelligence (hereinafter referred to as “AI”) added another knot to this already complicated affair.

The intersection of ghost-writing with AI has brought about a plethora of critiques concerning various disparities. Today, a ghost-writer appointed to draft a celebrity memoir uses AI tools to accelerate the process. Here, the rights of the employer, the ghost-writer, the AI tools, and the coder of the AI algorithm have vague horizons, leading to overall ambiguity. This study, therefore, aims to explore the nuances of ghost-writing in the age of AI while unpacking various legal, ethical, and IPR implications pertaining to it.

UNDERSTANDING THE CONCEPT OF ‘GHOST AUTHOR’

Traditionally, a ghost author is referred to situations where a person has contributed immensely 3 Aleksandra Nowak-Gruca, Could an Artificial Intelligence be a Ghostwriter?, 27 J. INTELLECT. PROP. RIGHTS

25 (2022).

Sandeep B.

Bavdekar, Authorship Issues,

LUNG

INDIA

(2012), https://journals.lww.com/lungindia/fulltext/2012/29010/authorship_issues.20.aspx (last visited Sep 30, 2025).

to an academic or literary work but remains uncredited. This is an unethical practice where the person who is contributing does not get recognition. This can be seen in the pharma industry, where companies hire professional medical writers to draft papers for publication and do not disclose them in the author lists. 5 This concept is the gap between labor and recognition, where the creator remains invisible behind a pseudo-author. This in itself is a grey area towards ownership of the content created, but when AI comes into the picture, it reshapes the notion by introducing a ghost author, which is not human. When AI produces content for a literary, painting, or musical composition, it acts as the de facto creator, but the recognition inevitably goes to the user or programmer. This creates a dichotomy with AI being the de facto creator versus the human who is the de jure author.

On one side, AI’s algorithms, along with vast learning and training data, produce the content, but on the other side, the contemporary copyright laws state that human involvement is required for a valid copyright. For example, in the United States, copyright law developed from the Patents and Copyrights clause of the Constitution. Later, in the Copyright Act of 1976, a work must have a human author to be eligible for copyright. 6 Additionally, Section 9(3) of the Copyright, Design, and Patents Act 19887 implies that the person who made the arrangements is the author. 8 This leads to the question of whether the output generated by AI should be left in the public domain or whether the creation should be rewarded to a human being.

COPYRIGHT FRAMEWORK

INDIAN FRAMEWORK

The copyright law in India recognizes human authorship. The legislature under the Copyright Act 1957 protects works only created by an author based on originality and rooted in the efforts made by humans. The Act provides for computer programs, which are listed as literary works, 5 Peter C. Gøtzsche et al., What Should Be Done to Tackle Ghostwriting in the Medical Literature?, 6 PLoS Med. e23 (2009), https://doi.org/10.1371/journal.pmed.1000023. 6 Matt Blaszczyk, Geoffrey McGovern & Karlyn D. Stanley, Artificial Intelligence Impacts on Copyright Law, RAND (2024), https://www.rand.org/pubs/perspectives/PEA3243-1.html#:~:text=Generative AI systems can produce, human.

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

8 Simon Stokes, WIPO Impact of Artificial Intelligence on IP Policy Response from Brunel University London, Law School & Centre for Artificial Intelligence About Brunel Law School About Brunel Centre for Artificial Intelligence About the Authors, 1 (2019).

but the output generated by the AI is not a program and hence falls outside the purview. Despite this, Indian courts have devolved into the aspect of AI being the author. In the case of Raghav AI v. Union of India,9 an AI unconventionally generated a digital painting. The Copyright Office initially granted the registration of the copyright, but later withdrew upon realizing that the work lacked human involvement. The Delhi High Court held that, as under Section 2(d)10 of the Act, an author means any person who creates the work, which means an AI cannot be quantified as an author. It held that works which lack human involvement cannot enjoy copyright protection. At present, India remains of the view that content generated purely by AI is not copyrightable. However, if the work involves human skills, talent, or judgment, it can be considered copyrightable.

COMPARATIVE LEGAL FRAMEWORK

The global copyright law reveals a vast uniformity in the commitment towards human authorship. The United States, the European Union, and India strictly adhere to the notion of required human involvement to grant copyrights. Several courts and copyright officers have consistently refused the registration of a work that is purely AI-generated. Upon a closer look, it reveals that different countries interpret copyright law in unique ways. The EU inherits and follows the Berne Convention, which places importance on the originality of creations that are genuinely intellectual works of human authors. In the US, the copyright office requires works to have been created by humans. 11 Meanwhile, in the UK, a more flexible approach is opted for by keeping technological and protecting creations that involve the assistance of AI.

12 From an international stance, the Berne Convention and the Trade-Related Aspects of Intellectual Property Rights (TRIPS) reinforce the aspect of human authorship. The convention’s persistent use of the term ‘author’ in reference to the creator of works and the recipient of protection, the term author implies only natural persons despite it not being defined 9 Ananya Thakur “Legal Implications of AI-Generated Works in Copyright Law: An Analysis of Raghav Artificial Intelligence v Union of India”. LawfulLegal. 30 June 2025. https://lawfullegal. in/legal-implications-of-aigenerated-works-in-copyright-law-an-analysis-of-raghav-artificial-intelligence-v-union-of-india/.

10 The Copyright Act, 1957, § 2(d), No. 14, Act of Parliament, (India).

11 U.S. COPYRIGHT OFFICE, COMPENDIUM OF U.S. COPYRIGHT OFFICE PRACTICES § 313.2 (3d ed.

2021).

12 Supra note 5.

within the Convention.13 The Berne Convention provides for author and their works, which implies only natural persons, and does not include AI, but it does not explicitly provide for it, and TRIPS reincorporates Berne Convention obligations.14 World Intellectual Property Organisation (WIPO) acknowledges that the copyright law is lucid when it comes to AIgenerated works, but has not taken any steps to broaden the scope of the author to include AI.15

ETHICAL AND POLICY CONCERNS

A renowned novelist and journalist, Jay Caspian King, once wrote that the desire to see our

own experiences through the lens of an artist or witness the creative accomplishments of others is what drives the act of reading, listening, or visualizing creative work. 16 All acts of creation are, in their forms, a demonstration that embodies the psyche, dexterity, and feeling of originality. Thus, art, music, literature, or any other entertainment sources are considered worth protection under the category of copyrights, considering the efforts and ingenuity of their production. However, the concept of ghost-writing has occupied a controversial yet operational position in creative and professional life since its emergence. Even though this practice has been ethically complicated since the outset, confidentiality agreements coupled with the need for professional refinement have bridged the gap.

Moreover, a new phase of ghost-writing was entered with the rise of AI, bringing about the distinction of there always being a ‘human’ as a ghostwriter and then there being an ‘algorithm’ as a ghostwriter. Today, contemporary AI is capable of generating coherent, tailored, and contextually nuanced content, reformulating content creation practices. Human ghostwriters are increasingly relying on these tools to accelerate the process or to bring about the desired outcome completely. However, the employment of AI has brought about a drastic swerve in the originality and the transparency of numerous works, therefore, misleading the reader gravely, thus providing interesting food for thought. The ethical breach, here, is approaching a masquerade to create work and pretending 13 Johannes Fritz, The Notion of ‘Authorship’ under EU Law—Who Can Be an Author and What Makes One an Author?

An Analysis of the Legislative Framework and Case Law, 19 J. INTELLECT. PROP. LAW PRACT. 552 (2024), https://academic. oup. com/jiplp/article/19/7/552/7614897? utm_source=chatgpt. com&login=false. 14 P. Bernt Hugenholtz & João Pedro Quintais, Copyright and Artificial Creation: Does EU Copyright Law Protect AI-Assisted Output? , 52 IIC INT. REV. INTELLECT. PROP. COMPET. LAW 1190 (2021).

15 Andres Guadamuz, Artificial I ntelligence and C opyright, WIPO MAGAZINE (2017),

https://www.wipo.int/web/wipo-magazine/articles/artificial-intelligence-and-copyright-40141(last visited Sep 30, 2025). 16 Jay Caspian King, What’s the Point of Reading Writing by Humans? | The New Yorker, THE NEW YORKER 100 (2023), https://www.newyorker.com/news/our-columnists/whats-the-point-of-reading-writing-by-humans (last visited Sep 30, 2025).

it to be the original work of the employer of the ghostwriter. 17 In the book Guide to Automated Journalism18, the author clearly states the significance of AI authorship declarations for transparency, considering how uncommon it is for AI authors to be formally acknowledged. 19 Today, the evolution of AI has reached an extent where it is capable of producing tangible dance patterns faithful to the music's character. 20 Nevertheless, it needs to be understood that AI does not ‘create’ traditionally; rather, it reassembles from vast databases, giving out outputs that imitate originality that does not stem from actual lived experiences. The assistance of AI for any work, along with the absence of acknowledgment, ends up misleading the reader about the origin of the research and the extent to which there has been human contribution.

Moreover, along with complicating the authorship declarations, copyright and IPR become a conflicting legal issue here. As far as the policies pertaining to such are concerned, countries like the United States, the European Union, and India explicitly recognize the need for protecting the work of a human by way of copyright. However, the legislation has not reached a stage that recognizes AI as an absolute author. It is considered that human skills, coupled with a sense of judgment, are the foundational basis for determining the need to copyright any work. Moreover, with the practice of AI being inculcated in almost every sector, the possibility of corporations that own these AI systems taking over ownership of everything is a threat. This adds to the question of accountability, considering there is no concrete legislation providing guidance for the same.

Therefore, it can be understood that the introduction of AI as an author has brought about various ethical and policy concerns that remain unaddressed, pertaining to the lack of legislation.

THE DILEMMA OF OWNERSHIP

The Question of who should own the AI-generated content lies at the center of the debate. Before examining the various approaches, it is crucial to understand that this issue arises from the absence of universal acceptance of identity for AI. Several approaches have been devised to tackle this question, each with both sides of the coin. These can be categorized into three Cheryl Conner, Is Ghostwriting Ethical?,

FORBES

(2014), https://www.forbes.com/sites/cherylsnappconner/2014/03/13/is-ghostwriting-ethical/ (last visited Sep 30, 2025). 18 Andreas Graefe, Guide to Automated Journalism, (Tow Center for Digital Journalism, Columbia Journalism School, 2016). 19 Fiona Draxler et al., The AI Ghostwriter Effect: When Users do not Perceive Ownership of AI-Generated Text but Self-Declare as Authors,

ACM

TRANS.

COMPUT.

INTERACT.

(2024), https://dl.acm.org/doi/pdf/10.1145/3637875 (last visited Sep 30, 2025).

20 Ioan-Radu Motoarca, AI, Copyright, and Pseudo Art, 26 YALE J.L. & TECH. 430 (2024).

approaches. Firstly, the programmer of the AI is represented as the author, and the copyrights are vested in him. This approach addresses that the outputs generated via AI amount to an extension of the programmer's original code, making them the authors of the content generated. But this approach faces impediments when AI databases are trained on vast datasets from multiple sources, making it arduous to identify specific human contributions to any given output. Secondly, the use of AI is represented by the author, and the copyrights are vested with him. The user of AI can claim authorship on the grounds of creative input prompts provided by them to arrive at the content generated. It is opposed that a simple prompt by the user may not sufficiently constitute a creative contribution towards the content generated to justify the claim for copyrights.

Thirdly, where AI itself is represented as the author, which is indeed highly controversial and deemed undesirable by the majority of legal systems. This would lead to granting AI a legal personality, therefore, primarily challenging the overall copyright legal framework. And lastly, where a corporation is represented as the author. Traditionally, the employer owns the employee's work during the course of employment. A company that is involved in developing an AI tool could claim authorship over any content generated by its AI. The company has invented the AI tool and has invested in its creation, so it shall reap the benefits that arise from it. To further complicate the issue, economic rights, which include distribution, reproduction, derivative work, and performance, present practical complications.

These rights allow human owners of a copyrighted work to enter into commercial agreements and decisions, including licensing agreements. Further moral rights, which include integrity, attribution, etc, also raise questions. Traditionally, moral rights protect the personal relationship between the creator and their works, which also tends to reflect the creator's personality and reputation rather than just recognition for the work. As under Article 6bi of the Berne Convention,21 moral rights, which are rights that the author possesses even after the transfer of the economic rights, which are limited to humans, are recognized, as they would be maintained after the death of the author until the expiry of the economic rights. This does not include AI within the scope of this

21 Convention for the protection of literary and artistic works, 1886, Art. 6bis.

provision. It is well known that AI software does not possess any consciousness, interest, dignity, or reputation; hence, rewarding them with moral rights would be impractical and impossible. For an AI to be awarded moral rights, it must show aspects of decision-making and reasoning, which leads to further questions, like if an AI system tends to show any value-based decisions, then would it still be termed as “Artificial” Intelligence? In the case of Thaler v. Perlmutter22, it can be determined that there is a consistent decline in the recognition of AI authorship claims. Courts have, in Zarya of the Dawn,23 established the difficulties of Human-AI collaboration, where it must be determined that sufficient human involvement is required to claim copyright protection. In Raghav AI v.

UOI, the court held that AI-generated work with minimal human involvement cannot be rewarded with protection under copyright; however, if the human imposes a sufficient amount of creative control on the collaborative content, then it can claim copyright. 24 In the ANI Media Pvt Ltd vs. OpenAI Inc, which is an ongoing case in relation to OpenAI taking ANI Media's copyrighted content to train the AI. These cases establish the incapability of the current framework to cope with the contemporary Human-AI authorships. 25

WAY FORWARD

Today, AI is omnipotent. The central issue concerns whether and to what extent AI-generated works can be attributed to a human author, especially pertaining to copyright law, and how this affects the moral, ethical, and legal dimensions of authorship and originality. With the introduction of AI, mandating the requirement for its disclosure can help reduce ambiguity and bring forth a greater level of transparency. Moreover, the introduction of AI-specific guidelines by the legislature, particularly for IPR, would lead to consistency in its employment and establish a foundational basis. Further, encouraging recognition of human-AI collaboration would prove to be an initiative to eradicate this vagueness and assist in recognizing the

22 Atreya Mathur, Case Review: Thaler v. Perlmutter (2023), CENTER FOR ART LAW (2023).

23 Analla Tony, Zarya of the Dawn: How AI is Changing the Landscape of Copyright Protection - Harvard Journal of Law & Technology, (2023), https://jolt.law.harvard.edu/digest/zarya-of-the-dawn-how-ai-is-changing-thelandscape-of-copyright-protection (last visited Sep 23, 2025).

24 Supra note 7.

25 Kotni Partner, Swagita Pandey Associate & Anushka Tripathi Intern, COPYRIGHT IN THE AGE OF AI: A

CASE COMMENT ON ANI VS . OPENAI , CS ( COMM ) 1028 / 2024 (2024).

contributions exclusively. Eventually, the value of AI lies in its method of employment, since it is both an opportunity and a jeopardy.