Updates on Elsevier policies about ethics and the use of Artificial Intelligence

Dr. Rebeca Aladro, Elsevier Publisher of the Astronomy portfolio Library and Information Services in Astronomy, 10 November 3-7, 2025, ESO Santiago de Chile Updates on Elsevier Ethics Policies And The Use of Artificial Intelligence Key concerns related to the use of Generative AI in scientific writing and publishing •Inaccurate information and unreliable sources •Generative AI tools may generate incorrect or made-up information such as citing non-existent references, making up authors, or suggesting non-existing reviewers. They may give incomplete or misleading responses and spread misinformation. •Ethical concerns and biases •The models often lack transparency. A GenAI tool may be based on biased training data and potentially infringe ethical standards including plagiarism, using copyrighted information without consent or proper attribution, or violating confidentiality and/or data privacy rights. •Environmental impact •Training and running an AI system, as well as using the applications they're built for, requires a great deal of computing power, resulting in increased carbon emissions, energy and water usage. 2 Key concerns related to the use of Generative AI in scientific writing and publishing (continued) •Lack of up-to-date information •When a GenAI tool is used of which the dataset is outdated it does not provide current news or information. The recommendations the tool provides are only as current as the knowledge cut-off date. •Not trained with scientific data •A limitation of consumerfocused GenAI tools such as ChatGPT is that they are not trained with scientific data. •Of course, there are also LLMs specifically designed for scientists, including many domain-specific models. Researchers have been using them for years. We expect that this trend will continue to grow. 3 •Homogenization of content •By optimizing for efficiency, GenAI systems can potentially lead to content becoming increasingly uniform and repetitive. Elsevier’s policies for authors, editors and reviewers on Generative AI •Use Gen AI to improve readability, grammar and language of work •Apply human oversight and control •Disclose other uses of Generative AI •Not list or cite Generative AI and AI-assisted technologies as (co) author 4 Authors Editors and reviewers Figures, images, artwork Please note the author policy only refers to the use of Generative AI in the writing process, and not to the use of AI tools to analyze and draw insights from data as part of the research process. Where to find Elsevier's GenAI Policies for Journals: •Generative AI policies for journals •Further guidance can be found in the Elsevier Responsible AI Principles Elsevier’s policies for authors, editors and reviewers on Generative AI •Not upload the manuscript into an AI toolthis may violate confidentiality and author’s rights •Not upload peer review report or editorial decision letters – they may contain confidential information as well •Generative AI should not be used to assist in the review, evaluation or decision-making process 5 Authors Editors and reviewers Figures, images, artwork Please note the author policy only refers to the use of Generative AI in the writing process, and not to the use of AI tools to analyze and draw insights from data as part of the research process. Where to find Elsevier's GenAI Policies for Journals: •Generative AI policies for journals •Further guidance can be found in the Elsevier Responsible AI Principles 8 Check Integrity Currently screens submissions for unauthorized authorship changes after submission. Future integrity signals are in development Revised manuscripts are compared against the original submission and changes to the author list are flagged in Check Integrity for review and validation by a Publishing Ethics Screening Analysts. Screening Analysts assesses and tracks the manuscript, and takes an action (i.e., reject, recommendation to editor, escalation for further investigation) Check Integrity is a manuscript submission screening tool used by a team of Publishing Ethics Screening Analysts to identify potential breaches of standards and policies that may compromise research integrity and publishing ethics. For Editors - iThenticate: Plagiarism detection - Proofig: Image manipulation - Simultaneous submissions - Duplicate submissions - Fake authors, paper mills - Authorship changes - Topic relevance - Find reviewers Research Integrity tools integrated on the Editorial system For Authors - Find the correct journal - Prepopulating manuscript information - Transfer options - Checking OA agreements What is Scopus AI and how does it help researchers? *Interface and functionality subject to change. 12 1 2 3 4 Natural language queries Ease your search: Ask questions about a subject in a natural, conversational manner. Summary with Scopus references Instant overview: Skip the lengthy reading. Read a concise and trustworthy summary with academic references for each search. Dig deeper with expanded summaries. Visual representation of entities See the big picture: View a graphical representation of the keywords to reveal hidden connections and insights. Foundational papers Discover influential papers. Rapidly pinpoint seminal works, navigating academic progress and impact with precision and ease. 2 1 5Academic Expert Search See who the experts are. Rapidly identify the leading authors based on your query alongside explanations of their relevant expertise. 6Deeper query exploration Uncover more: Explore beyond the surface. Our AI offers relevant queries to discover new perspectives. 3 4 5 6