GENERATIVE AI

GENERATIVE AI POLICY

For Authors

The application of AI-assisted technologies and generative AI in scientific writing

The use of AI technologies to evaluate and derive insights from data as part of the research process is not covered by this policy; it solely relates to the writing process. 

Generative AI and AI-assisted technologies should only be utilized by authors to enhance the work's language and readability.  Since AI can produce output that seems authoritative but may be inaccurate, prejudiced, or incomplete, authors should carefully analyze and adjust the results before implementing the technology.  The authors are ultimately in charge of and answerable for the work's contents.

When using AI and AI-assisted technologies, authors should disclose this information in their manuscript; a statement will be included in the final product. By disclosing the usage of these technologies, authors, readers, reviewers, editors, and contributors can benefit from increased transparency and confidence as well as easier adherence to the appropriate tool or technology's terms of use.

It is improper for authors to cite AI as an author or name AI and AI-assisted technologies as co-authors.  Authorship entails duties and obligations that are exclusive to and carried out by humans.  Authorship necessitates the ability to approve the final version of the work and consent to its submission. Each (co-)author is responsible for making sure that any concerns regarding the precision or integrity of any portion of the work are suitably looked into and addressed.  Before submitting, authors should read our Ethics in Publishing policy to make sure their work is original, fits the requirements for authorship, and does not violate the rights of other parties.

 

Using AI-assisted tools and generative AI to create figures, pictures, and artwork

Generative AI and AI-assisted techniques are not allowed to be used to create or modify images in submitted publications.  A feature within an image or figure may be enhanced, obscured, moved, eliminated, or introduced.  Changes to brightness, contrast, or color balance are permissible provided that they don't obfuscate or remove any information from the original.  To find possible image anomalies, submitted articles may be subjected to image forensics techniques or specialist software.

The only exception is if the research design or procedures include the use of AI or AI-assisted tools (e.g., AI-assisted imaging ways to create or interpret the underlying research data, for example in the field of biomedical imaging).  The methods section must include a replicable description of the use if this is carried out.  A description of the model or tool's name, version and extension numbers, and maker should be included, along with an explanation of how AI or AI-assisted tools were employed in the image generation or change process. In order to ensure proper content attribution, authors must follow the particular usage guidelines of the AI software.  For editorial evaluation, authors may be requested to share pre-AI-adjusted photos or the composite raw images that were utilized to produce the final submitted versions, if relevant.

It is not allowed to create artwork, such as graphical abstracts, using generative AI or AI-assisted techniques.  In some circumstances, the author may be permitted to use generative AI to create cover art as long as they have prior approval from the publisher and journal editor, can show that all required rights have been obtained for the use of the pertinent content, and make sure that proper content attribution is made.

 

For Reviewers

Implementing AI-assisted and generative AI in the peer review process for journals
A researcher must consider a manuscript as a confidential document when they are asked to examine another researcher's work.  A submitted manuscript, or any portion of it, should not be uploaded by reviewers to a generative AI tool because doing so could infringe upon the authors' proprietary and confidentiality rights and, in cases where the work includes personally identifiable information, violate data privacy rights. 

Since the peer review report may include private information about the authors or the article, it is also subject to this confidentiality requirement.  For this reason, even if reviewers are only using an AI tool to make their peer review report more readable and linguistically correct, they should not upload it.

The foundation of the scientific ecosystem is peer review, and Elsevier upholds the highest standards of honesty in this procedure.  The act of reviewing a scientific publication has obligations that are unique to humans.  Reviewers should not use generative AI or AI-assisted technologies to help with the scientific review of a paper because, in addition to being unable to perform the critical thinking and original assessment required for peer review, these technologies run the risk of producing biased, inaccurate, or incomplete conclusions about the manuscript.

According to SUKUK's AI author policy, writers are permitted to employ generative AI and AI-assisted technologies throughout the writing process prior to submission, provided that they provide the necessary disclosure and enhance the paper's language and readability in accordance with our guidelines. 

SUKUK welcomes innovative AI-powered tools that assist editors and reviewers in the editing process while upholding the rights to data privacy and confidentiality of authors, reviewers, and editors.

 

For Editors

 

The implementation of AI-assisted technologies and generative AI to the journal editing process

A manuscript that has been submitted needs to be handled as a private document.  Editors should refrain from uploading a submitted manuscript, or any portion of it, to a generative AI tool as doing so could infringe upon the authors' proprietary and confidentiality rights and, in cases where the paper includes personally identifiable information, violate data privacy rights. 

Since notification or decision letters may contain sensitive information regarding the article and/or the authors, they are also subject to this confidentiality requirement.  Therefore, even if editors want to improve the language and readability of their letters, they shouldn't upload them to an AI program.

In the process of peer review, which is vital to the scientific ecosystem, SUKUK upholds the highest standards of honesty.  The responsibility of overseeing the editorial review of a scientific publication is exclusive to humans.  Editors should avoid using generative AI or AI-assisted technologies to help with manuscript evaluation or decision-making because, in addition to being unable to perform the critical thinking and original assessment required for this work, these technologies run the risk of producing inaccurate, partial, or biased results regarding the manuscript.  The editorial process, the ultimate decision, and informing the authors of it are all under the editor's control and accountability.