AI in academic writing: Transparency and ethics

To the Editor,

I read with great interest the recent article by Durrani et al.,[1] titled “Evolution of artificial intelligence writing tools in THE domain of scientific writing: A threat to human writing and its ethical considerations,” published in the Journal of Family Medicine and Primary Care (2025). The paper is aptly timed, with a surge of multitude of artificial intelligence (AI) tools, originating every day. The article thoroughly examines the impact and ethical aspects related to AI-assisted scientific writing tools.

The authors have traced the journey and development in the field of AI writing technologies from early grammar checking programs like “Grammarly” to more sophisticated large language models like ChatGPT and Google Bard, which have revolutionized the way researchers write manuscripts. The authors have critically discussed the dual role of these tools, which is increasing the writer’s productivity and language accuracy on one hand, while also compromising the originality and critical thinking of the writers in the long run. The authors also acknowledge that though AI can help in routine tasks like grammar correction, citation generation, and summarization, it cannot yet replace the multidimensional creativity, complex comprehension, and the intellectual ability, which is needed for high-quality scientific writing.

Additionally, various AI-powered image generation tools are increasingly being utilized in academic research and publications.[2] These tools can create figures, diagrams, and visual abstracts, which provide a clearer and easier understanding of complex scientific details. However, the use of this tool also raises ethical concerns regarding authenticity and intellectual property rights, which the authors should carefully analyze before using these tools.

Durrani et al. emphasize the ethical challenges associated with the use of AI in academics, including the biases which might have occurred during the training of these tools, concerns about the privacy of the inputs as well as the output, and also the difficulty in distinguishing AI-generated content from human work. The author correctly advocates for clear guidelines and licensing to allow use of AI tools in academic work, so that these technologies can be utilized by all, so as to utilize these tools to help the academicians, and omit the concerns that these tools might replace the work of mankind.

Academicians are asked to use AI responsibly by verifying AI-generated content rigorously, and also to acknowledge the use of AI tools transparently. The International Committee of Medical Journal Editors strongly recommends the authors to disclose any AI assistance in manuscript preparation. Additionally, entities like the World Association of Medical Editors and Committee on Publication Ethics clarify that AI tools cannot be credited as authors, as AI cannot be made accountable for any published work.[3]

In conclusion, Durrani et al.’s work is a great contribution at the apt time, and promotes a balanced approach for the integration of AI in scientific writing. Transparent discourse by the author and policy development in this area is the need of the hour, with increasing integration of AI tools in the academic ecosystem.

Author contributions

Seema Shah contributes in conceptualization, methodology, validation, formal analysis, investigation, data curation, and writing – original draft. Neha Rani Verma contributes in resources, conceptualization, supervision, and writing – review and editing. Eli Mohapatra contributes in conceptualization, supervision, and editing the final draft.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

Nil.