Acciones e investigaciones Sociales. Nº 47 (2026)

Esta obra está bajo una licencia internacional Creative Commons Atribución-NoComercial 4.0. 285

Can I co-author with my robot? The legal, ethical and economic dilemmas of

doing academic research in times of artificial intelligence

¿Puedo escribir un artículo con mi robot? Los dilemas legales, éticos y

Received 2025-05-18. Approved 2025-12-09

DOI: https://doi.org/10.26754/ojs_ais/ais.20254711876

Abstract

Currently, there is a huge wave of enthusiasm in the education sector about artificial intelligence

which materializes on the great number of documents and courses to capacitate teachers and

students in mastering AI apps. However, it is possible to find some reservations among the sector,

particularly in the publishing industry and universities that manage scholarly journals. These doubts

incorporation of AI for academic research. Final outcomes will show that these contradictory views on

AI are also evidence of the deep impact of this disruptive technology in the academic publishing sector,

and how legal frameworks and financing mechanisms are becoming outdated to comply with the quick

changes introduced in the educational sector.

Keywords. Academic Research; Ethics and AI; Intellectual Property; Authorship.

Acciones e investigaciones Sociales. Nº 47 (2026)

Esta obra está bajo una licencia internacional Creative Commons Atribución-NoComercial 4.0. 286

Resumen

Actualmente, existe un gran entusiasmo en el sector educativo por la inteligencia artificial que se

materializa en el número de documentos y cursos para capacitar a profesores y estudiantes en el

del plagio, sino también por cuestiones económicas y legales. Este análisis se centra en la revisión de

la literatura académica, pero también en el estudio comparativo de textos legales (regulaciones y

sentencias), así como el examen de casos judiciales controvertidos sobre la incorporación de la IA para

la investigación académica. Los resultados finales mostrarán que estas posturas contradictorias sobre

la IA también son evidencia del profundo impacto de esta tecnología disruptiva en el sector editorial

académico, y cómo los marcos legales y los mecanismos de financiamiento se están volviendo

obsoletos para hacer frente a los rápidos cambios introducidos en el sector educativo.

Palabras clave: Investigación Académica; Ética e IA; Propiedad Intelectual; Autoría.

INTRODUCTION

privacy, the substitution of humans by robots in diverse economic sectors, and the latent menace of

security outbreaks. Regarding education, there is a huge wave of enthusiasm that materializes on the

number of documents and courses to capacitate teachers and students in mastering AI apps.

Educational institutions have not been hesitant to introduce AI applications into the classroom to

support the learning process by helping teachers to improve their lesson plans and find bibliographies,

and assist students to translate, summarize and edit homework. However, it is important to mention

some critical voices that have accentuated the potential learning setbacks by leaving the whole

student education process in hands of AI systems without ethical and pedagogic guidelines (Kosmyna

et al. 2025; Veale et al. 2025).

Acciones e investigaciones Sociales. Nº 47 (2026)

Esta obra está bajo una licencia internacional Creative Commons Atribución-NoComercial 4.0. 287

Researchers have also introduced AI to assist them while conducting academic investigations:

translating articles, summarizing data, supporting writing, and proofreading. However, it is possible to

find some reservations among the sector, particularly in the publishing corporations and universities

enthusiasm vs reservations, are motivated, not only to preserve academic ethics and limit the

proliferation of malpractices as the case of plagiarism, which is claiming other authors’ ideas and/or

intellectual work as their own (Authors Alliance, 2020), but what these contradictory views also

evidence is the huge impact of disruptive technology in the academic publishing sector, and how the

legal framework and financing mechanism are becoming outdated to comply with the quick changes

introduced by AI in the sector.

AI AND ACADEMIA

there have been studies on the use of AI apps to support scholars. This assistance usually translates in

the automation of repetitive activities, such as the data processing and the discovery of patterns and

categories (Romero, 2023). Regarding the process of composing the document, it is possible to

mention some apps such as Smartpaper.AI. According to the developer “Smartpaper.ai is an artificial

intelligence tool designed to facilitate the entire process of writing a scientific/academic research

project” (https://smartpaper.ai/about/). While there is great enthusiasm about the benefits of using

this application to create, from articles and chapters to doctorate theses, there are also important

concerns regarding its use (Romero, 2023). Some of these apprehensions could be appreciated even

in the app’s official website. In the section “frequent questions” (https://smartpaper.ai/faqs/), it is

possible to read some topics about plagiarism and the ethical implications of this software

Acciones e investigaciones Sociales. Nº 47 (2026)

Esta obra está bajo una licencia internacional Creative Commons Atribución-NoComercial 4.0. 288

academic work was created using Smartpaper. Another similar app, and by far the most used, is

ChatGPT. Launched by the end of 2022, this chatbot, based on GenAI, enjoyed great popularity among

students, teachers and researchers. The easy-to-use characteristic of ChatGPT contrasts with its great

aspects of the software but also to recall the ethical responsibility of the authors, who, at the end,

make the decisions about how to employ the application (Padilla-Caballero et al., 2023).

Besides the popularization of AI in academic research, there are still conflicting positions. In 2024,

Camino & Clavijo (2024) made a survey among Ecuadorian teachers, while 40.74% declared that they

have employed AI to compose a document, when asked about considering co-authorship with AI,

81.48% of the respondents consider that it is important to set a verification system that detects the

use of AI in the composition of academic works. Other research, centered on the Colombian context,

showed the same contradictory feelings about the employment of AI at school. Thus, Díaz-Cuevas &

ARTIFICIAL INTELLIGENCE AND THE PUBLISHERS’ DILEMMA

Besides the great eagerness regarding the introduction of AI to improve education and research, this

fervor has not infused deeply among publishing companies (Lund & Naheem, 2024). This is not only

about publishers that sell non-fiction books and journals, but also among editors that manage free-

open journals and editorial series. While some companies have recently signed agreements with AI

developers to let their models be trained using their content, as the case of Taylor & Francis and

Microsoft (Kwon, 2024), most of them still have some reservations regarding the inclusion of AI as

authors. Big publishing corporations such as Springer Nature (Springer, 2025) have declared that, by

now, “Large Language Models (LLMs), such as ChatGPT do not currently satisfy our authorship

criteria.” This, of course, is a decision that not only stands on the scientific principles, such as data

Acciones e investigaciones Sociales. Nº 47 (2026)

Esta obra está bajo una licencia internacional Creative Commons Atribución-NoComercial 4.0. 289

infringement. But it also refers to a question of economic nature, since there are worries if these

contents (books, articles or book chapters) are susceptible of being marketed.

In the last two years, other institutions have followed similar steps. In 2023, H. Holden Thorp Editor-

policies to stand that "Nonhuman artificial intelligence, language models, machine learning, or similar

technologies do not qualify for authorship." (Flanagin et al., 2023, p. 03). JAMA’s editors adduced trust

and accuracy issues to set these prohibitions. AI applications such as ChatGPT provide responses that

sometimes are "not up to date, false or fabricated, without accurate or complete references, and

worse, with concocted nonexistent evidence for claims or statements it makes" (Flanagin et al., 2023,

p. 03).

While other editors, such as William J. Jr. Dupps from the Journal of Cataract & Refractive Surgery,

praises the capabilities of artificial intelligence," AI has enormous potential to enhance and accelerate

respect the efforts of every author that has collaborated in the composition of the work. The second

issue focuses on the legal aspects of intellectual property rights. This is not only about the economic

aspect, but also on the moral rights. That is to be credited as the author, far beyond the commercial

operation of the work. The final issue mainly links to preserving the accuracy of the information and

following the scientific research criteria. However, it is important to mention that besides the

examples cited above, editorials’ restrictions about AI tend to vary, depending on the journal’s

discipline, from a total constraint up to the obligation of mentioning which parts were created using

AI. But what almost all journals agree on (98.9% of 300 journals) is that AI cannot be enlisted as an

author (Lund & Naheem, 2024, p. 14). The following section will focus on the problem of academic

authorship and artificial intelligence, particularly on the economic, legal and ethical arguments about

Acciones e investigaciones Sociales. Nº 47 (2026)

Esta obra está bajo una licencia internacional Creative Commons Atribución-NoComercial 4.0. 290

DIGITALIZATION, INTELLECTUAL PROPERTY AND ACADEMIC PUBLISHING

As digital media, from smartphones to social platforms and mobile applications, globally consolidates

as an essential element in people’s daily lives, and most all the content and information can be

smartphones, as a basic legal instrument to encourage innovation and creativity, but as digital media

evolved and dispersed into the social, political and economic sphere, the traditional forms of

distribution and legal protection of the content have been compromised. For decades, and until the

advent of the Internet, the market distribution of books, newspapers, music and movies have been

based on the analogical format. Media corporations, distributors and copyright holders were able to

control and measure the potential revenues by counting the number of books, DVDs, newspapers and

CDs sold. However, with the digitalization of media content and the diversification of online

distribution channels, this traditional mechanism of supply has entered into a long crisis and decline.

not been immune to these effects brought by the Digital Era (Hviid et al., 2016). While the introduction

of computers and the Internet in the publishing industry has carried advantages: decentralization of

the organizational infrastructure and outsourcing of diverse processes to reduce production costs, the

transition to the digital format system by which corporations sell their content in big media platforms

has also boosted the piracy of copyrighted material, not just books, but tutorials, courses, and

academic articles that freely circulate online. In addition, the traditional model of academic publishing

has been criticized by the excessive price of the content, which contrasts with the lack of proper

retribution to authors, reviewers and editors, whose unpaid labor is an essential pillar that sustains

the entire academic publishing system. In recent years, it is becoming more frequent the questionings

about the honorary nature of these activities (Cheah & Piasecki, 2022). Additionally, there is a popular

Acciones e investigaciones Sociales. Nº 47 (2026)

Esta obra está bajo una licencia internacional Creative Commons Atribución-NoComercial 4.0. 291

& Souza, 2024; Esteve, 2024). Among the transition from analog to digital format, the arrival of AI

stands more challenges for the media and creative sector. The creative sector (radio, cinema, music

and publishing, etc.) has not been the exception.

Among the main issues was the establishment of limits to the employment of AI to create movie and

TV show scripts. According to the arrangement, the studios “cannot use AI to write scripts or to edit

scripts that have already been written by a writer” (Anguiano & Beckett, 2023). However, the

employment of generative AI is not absolutely forbidden as the Hollywood studios were allowed to

use apps such as ChatGPT to generate drafts that could be modified by the writers. This popularization

of Generative AI has compromised the economic situation of traditional media corporations which are

trying to survive the end of the analogical era. In the absence of physical formats, the copyright holders

stand on the legal protection of their content to distribute and generate profits, not only by selling

material free of charge. Their approach has changed to a more open and collaborative system,

particularly considering AI, in which they are willing to work with companies such as Microsoft to allow

AI systems to be nourished using copyrighted content. In addition, some of them have promoted their

open-access models to allow authors to share their works across the Internet. However, the

employment of AI to create content entails other complex challenges to the intellectual property

regulation. The next section will center on these issues, the authorship dilemma.

THE LEGAL AND ETHICAL QUESTIONS ABOUT CO-AUTHORING WITH AI

As it has happened with other rights that become compromised by the arrival of disruptive

technologies, such as privacy or free speech, intellectual property faces the impact of the

Acciones e investigaciones Sociales. Nº 47 (2026)

Esta obra está bajo una licencia internacional Creative Commons Atribución-NoComercial 4.0. 292

subject of legal and economic scrutiny since the capabilities of AI not only provide the human authors

with robust tools to create works, but it also makes decisions during the creative process (Guadamuz,

2017). Of course, this was not a problem when, one of the first legal instruments that protects

Literary and Artistic Works, 1886, Article 6bis). For decades, these principles functioned well as it

covered traditional relations among authors and the corporations or individuals that held the rights

to commercialize their works. However, applying these principles (A) Humans as authors and (B) the

right of preserving the integrity of the works, in the age of artificial intelligence, implies some barriers:

(1) the limitation that only humans can be authors. To change this, the regulation and, but above all

our legal tradition on what could be considered as creator, should be amended, and the second barrier

is (2) the preservation of the works, as applications such as ChatGPT, depart from copyrighted material

to deliver new content “inspired” by the original “output.” However, sometimes users employ AI to

court case about AI and authorship, a Czech court ruled that AI-generated works cannot be under the

protection of copyright law, as AI “is not a “natural person, it cannot be the creator of a work under

European copyright law” (Novagraaf Team, 2024). Next year, in March 2025, The Court of Appeals for

the District of Columbia (DC) circuit (USA) supported the rejection of a copyright application to protect

a work generated solely by AI. The court concluded that the work cannot be recognized as an author

under actual copyright regulation (Yi & Chau, 2025). The similar criterion was upheld by other courts,

as in the USA and EU. In 2024, The Specialized Chamber on Intellectual Property Matters of the Federal

Court of Administrative Justice (TFJA) in Mexico, supported the decision of the INDAUTOR (Federal

Copyright Office) of denying the protection of a work created by AI (Gómez Viñuela, 2025). The court

adduced that, according to the actual Mexican copyright laws, only “natural persons who create works

Acciones e investigaciones Sociales. Nº 47 (2026)

Esta obra está bajo una licencia internacional Creative Commons Atribución-NoComercial 4.0. 293

autores)(Ley Federal de Derechos de Autor, 1996, art. 12). Though, besides these legal precedents,

there have been two cases where Chinese courts have pronounced to defend the copyright of content

created using AI (Wininger, 2025). In 2025, The Changshu People’s Court protected the copyright of

regulations. Then it is important to go deeper on the concept of an author, and how only “natural,”

“physical” and “human” persons, depending on the composition of the legal framework, could be

considered as an author. In the Spanish regulation, an author is “considered to be the natural person

who creates a literary, artistic or scientific work.” (Ley de Propiedad Intelectual, 1996, art. 5), A similar

measure has been established in EU and USA regulations (Hutukka, 2023) and in Latin American legal

systems (Ávila Vallecillo, 2021). This anthropocentric criterion to determine authorship bases not only

on practical principles but also philosophical and even religious assumptions. When Alan Turing

proposed his “imitation game” to question if machines are able to exhibit intelligent behavior

considering the huge technological advances in computer science, is it still valid to maintain this

principle intact?

Considering the legal technical aspects regarding AI authorship, Morales Cáceres (2021) proposes

three main standards (a) To deny legal protection to works in where the human participation was null

or insignificant, (b) To protect the work if a natural person has participated on the inputs, providing

the AI system with the guidelines to create the content or (c) To adjudicate the authorship to the

copyright holder of the AI system. Other authors as Portalés (2022) proposed the creation of a

ciberhumanoid, that is to bestow juridic personality to an AI entity. While this proposal could look

futuristic and has been only portrayed in science fiction movies such as “Blade Runner” or “Her,” the

diversification, constant improvement and popularization of Generative AI applications challenge this

Acciones e investigaciones Sociales. Nº 47 (2026)

Esta obra está bajo una licencia internacional Creative Commons Atribución-NoComercial 4.0. 300

Knibbs, K. (2024, June 24). Discográficas demandan a los generadores música con IA Suno y Udio por

Ministerio de Cultura de España (1996). Ley de Propiedad Intelectual.

Gobierno de México (1996). Ley Federal de Derechos de Autor.

Lund, B. D., & Naheem, K. t. (2024). Can ChatGPT be an author? A study of artificial intelligence

Menell, P. S. (2023). Mapping the Intellectual Property/Social Justice Frontier. In S. D. Jamar & L.

Morales Cáceres, A. (2021). El impacto de la inteligencia artificial en el Derecho. Advocatus, 039,

Novagraaf Team. (2024, May 1). AI and copyright. Novagraaf.

Oddone, N., & Souza, L. V. R. L. de. (2024). Access to knowledge in the context of Open Science: