Harnessing Artificial Intelligence Responsibly: Advancing Scientific Publishing Amid New Challenges

Linda Trinh, PhD, FSBM - The Publication Council

Disclosure of use of Artificial Intelligence (AI)¹

During the preparation of this work, the authors used Chat GPT to generate a punchy and attention-grabbing title in hopes of captivating readers. Specifically, the prompt ‘come up with catchy titles with the keywords AI, scientific publishing, and challenges’ was entered. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the published article, including using large language models (LLMs).

LLMs are transforming scientific publishing, creating challenges for authors, researchers, and publishers. Oxford University Press (OUP) manages SBM’s flagship journals: Annals of Behavioral Medicine (ABM; Editor-in-Chief: Dr. Angela Bryan) and Translational Behavioural Medicine (TBM; Editor-in-Chief: Dr. Cheryl Knott). “As a leading publisher of high-quality scholarly journals, OUP is committed to maintaining the highest standards of publication ethics and to supporting ethical research practices” (OUP, 2025). OUP recognizes the increased use of LLM tools in academic research and the benefits to assist authors in the research and writing process. Such tools, where used appropriately and responsibly, have the potential to supplement research outputs (e.g., synthesizing, polishing language, analyzing findings).

Many academic publishers such as OUP have formulated LLM guidelines to assist authors, reviewers, and editors to make responsible, appropriate, and transparent judgements about the ethical use of such technology. In these guidelines, ‘AI’ refers to applications, tools, and programs using AI. These are not limited to large language models (LLMs), AI, and chatbots (OUP, 2025). For example, ChatGPT is a large language model from OpenAI that uses deep learning to generate human-like responses in natural language based on information requested by a user in a prompt (OUP, 2025).

The guidelines for the use of LLMs in research publishing are informed by three principles which are briefly outlined below (OUP, 2025).

1. Authorship

• LLMs do not qualify as an author and should not be used to undertake primary authorial responsibilities, such as generating arguments and scientific insights, writing analysis, or drawing conclusions.
• Authorship requires taking accountability and integrity for the work, consenting to publication via a publishing agreement, among other principles. These are uniquely human responsibilities that cannot be performed by LLM tools.

2. Accountability

• Authors remain fully accountable for the accuracy, integrity, and originality of their works, including any LLM-created content these works may contain.
• Authors are therefore liable for any breach of publication or research ethics from their use of LLM tools. This includes inaccuracy, so-called ‘hallucination’ (where the LLM tools simply makes things up, including plausible-sounding citations; lack of up-to-date knowledge), a lack of full attribution and referencing, inadvertent plagiarism, and bias.

3. Disclosure

• The corresponding author should indicate on the Title Page (not in the body of the manuscript), whether artificial intelligence was used in the development of the manuscript.  Options for this disclosure include: 
  • “Artificial intelligence was not used in the development of this manuscript.” OR 
  • "During the preparation of this work the author(s) used [NAME TOOL / SERVICE] in order to [REASON]. After using this tool/service, the author(s) reviewed and edited the content as needed and take(s) full responsibility for the content of the published article, including analysis of statistical data using AI tools." 

Furthermore, the impact of LLMs extends into the peer review process. Peer reviewers are chosen experts in the field or methods of work to evaluate content based on their expert judgement alone. Peer reviewers play a critical role in scientific publishing as their expert evaluations guide editors in their decisions. Peer reviewers are accountable for the accuracy and integrity of their submitted reviews to ensure that the research is valid, rigorous, and credible. Many academic publishers, including OUP, prohibit editors and peer reviewers from uploading grant proposals and manuscripts, in part or in whole, into  LLMs for any purpose (e.g., to generate review reports, article summaries, decision letters) (OUP, 2025; Taylor & Francis, 2025; Springer Open, 2025). Doing so may infringe on intellectual property, copyright, confidentiality, privacy, and data security obligations. The peer review process operates on a principle of mutual trust between authors, reviewers and editors to maintain confidentiality (OUP, 2025; Taylor & Francis, 2025; Springer Open, 2025).

Overall, navigating LLM policies presents challenges to authors. In general, academic publishers do not accept LLM-authored manuscripts, but they may permit submissions with some AI content to improve the manuscript (e.g. style, generating reference list, images, graphs). However, transparency and disclosures about the use of LLMs will remain a forefront of LLM policies for many academic publishers. While the use of LLMs in scientific publishing may be transformative, they are unable to replicate human wisdom and critical thinking (at least for the near future).

References:

1. Oxford University Press (November 2025). Preparing and submitting your manuscript. https://academic.oup.com/pages/for-authors/journals/preparing-and-submitting-your-manuscript/ethics?login=false#Authorship
2. Taylor and Francis (November 2025). AI policy. https://taylorandfrancis.com/our-policies/ai-policy/
3. Springer Open (November 2025). Editorial policies. https://www.springeropen.com/get-published/editorial-policies#artificial+intelligence+%28ai%29