JIBE statement on the use of artificial intelligence

December 2024

Introduction
This statement aims to provide greater transparency and guidance for JIBE authors, readers, reviewers, and editors in relation to the use of Artificial Intelligence (AI) tools.
AI tools, including Generative AI (GenAI) or Large Language Models (LLMs), continue to rapidly develop and evolve. These tools are useful for idea generation and exploration, linguistic refinement, figure enhancement, and where appropriate as data analysis tools.
JIBE recognizes the value of AI and welcomes the ethically responsible use of tools that assist with generating quality research outcomes and scholarly manuscripts.
While these tools present powerful opportunities for undertaking, communicating, visualizing, and disseminating research, it is important to remain mindful that AI are merely tools. AI is not able to replicate the unique aspects of genuine human creativity, critical thinking, meaning making, determining importance, and judge societal implications.
AI, particularly GenAI, can present unique risks that authors must be aware of. These include:
• Inaccuracy, bias, and overgeneralizations: AI tools are driven by algorithms and are statistical in nature, based on the data it draws intelligence from. AI also relies on content that continues to be added to, thus AI is continuously evolving and may result in AI-driven outcomes changing over time. As such, AI can possibly introduce inaccuracies, flaws, falsehoods, or bias that are difficult to detect, verify, and correct.
• Breaching confidentiality and intellectual property rights: At present, AI draws from data that could infringe, unknowingly to the authors, on confidential information, data security, intellectual property rights, and copyright protection.
• Lack of attribution: AI often lacks the ability to correctly and precisely attribute ideas, quotes, or citations to the original source, therefore, potentially introduce intellectual inaccuracy and dishonesty.
• Unintended additional uses: Where the tool is not secured, any submitted information can be used to further AI learning and may be provided to others. Therefore, the use of AI to analyze data may result in breach of human research ethics approval constraints (e.g., the common requirement that only authors have access to the raw data), institutional policies on data protection, data sovereignty, etc.
• Infringement on rights: AI platforms may reuse the information that authors have uploaded, this could potentially infringe on the rights of authors and publishers, the right for privacy of research participants, among others.

Responsibilities of the authors
It is fully the responsibility of the authors to ensure compliance with appropriate safeguards and protections (e.g., intellectual property rights, confidentiality, and security), human research ethics conditions, accuracy of the interpretation, and that outcomes meet the rigorous expectations of scholarly assessment. Statements that provide such assurance may be required in the manuscript.
Transparency of how research was conducted and how ideas were generated is an important principle in scholarly work. Therefore, where AI is used, the authors must state in the manuscript:
1. What AI tool that was used, including the version.
2. How it was used.
3. Why it was used.
JIBE does not allow the use of GenAI to create scholarly content, write script, or interpret and create meaning from the research outcomes. AI cannot be an author or be used as an author because AI is not able to take responsibility of the accuracy, reliability, and validity of the interpretation of the research outcomes and the importance of the scholarly meaning – these are uniquely human responsibilities.

Responsibilities of the editors and reviewers
JIBE strongly adheres to the principle that the manuscript review process is confidential and restricted to the editors, relevant reviewers, authors, and the publisher. AI, and in particular GenAI, may pose a risk to confidentiality and proprietary rights of information, may result in self-identifying information being shared, and present other risks. Therefore, editors and reviewers must not upload files, images, or information from unpublished manuscripts into GenAI tools or use these tools to analyze the manuscript. Editors and reviewers can use AI for linguistic improvements of the review recommendation text only if using a secure tool that is integrated into, for example, Microsoft Word, supported by the reviewer’s institution, and where it clearly states that it is a secure tool. Unsecure tools (e.g., ChatGPT) use submitted information for further learning and may result in parts of the text being reproduced and shared elsewhere, which breaches the review confidentiality.

Undisclosed or inappropriate use of AI
If undisclosed or inappropriate use of AI in a submitted manuscript or published article is suspected, JIBE will undertake an editorial assessment. This will include an investigation of the concerns raised, guided by JIBE policy and advice from the Committee of Publication Ethics (COPE). If this investigation finds that there has been undisclosed or inappropriate use of AI, this will be seen as unprofessional behavior and in breach of the JIBE publishing policy and may result in submitted manuscripts being rejected and published articles being retracted.