Written and maintained by CASRAI Editorial Board
Last updated
Elsevier requires authors to disclose any use of generative AI or AI-assisted technologies in preparing a manuscript, through a specific declaration statement that appears in the published article. This page gives the exact required wording, where it goes, what is exempt, and how Elsevier’s policy handles authorship, figures, and peer review.
Quick answer: what to write
Add a section titled “Declaration of generative AI and AI-assisted technologies in the manuscript preparation process” at the end of the manuscript, immediately before the reference list. Elsevier’s suggested wording is:
“During the preparation of this work the author(s) used [NAME OF 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 publication.”
Fill in the tool name and the specific reason it was used (for example, “to improve readability and language” or “to check grammar and phrasing”), then state that the author(s) reviewed the output and take full responsibility for it. If you used no generative AI tools in the writing process, you do not need to add this section at all.
Illustrative worked example
The following is an illustrative example showing the template filled in — not a statement from any real, named publication:
“During the preparation of this work the author(s) used a large language model (GPT-4) in order to improve the readability and language of an early draft. After using this tool, the author(s) reviewed and edited the content as needed and take full responsibility for the content of the publication.”
What does and doesn’t need disclosure
| Use | Disclosure required? |
|---|---|
| Drafting, paraphrasing or rewriting sentences with a generative AI tool (e.g. ChatGPT, Copilot) | Yes |
| Generating or restructuring an outline or section with an AI tool | Yes |
| Basic grammar, spelling and punctuation checks (including standard spell-checkers) | No — explicitly exempted |
| AI-assisted data collection or analysis methods (e.g. AI-assisted imaging) | Described in the Methods section, not the AI declaration |
Where the statement goes
Place the declaration at the end of the manuscript, immediately before the reference list — after acknowledgements and funding/competing-interest statements, not in the abstract or methods. Elsevier states the declaration will be preserved in the published version of the article, so authors should treat it as a permanent part of the record, not a submission-only formality.
AI tools cannot be listed as an author
Elsevier’s policy is explicit that generative AI tools, and AI-assisted technologies more broadly, cannot be credited as an author or co-author on a manuscript. The stated rationale is that authorship carries responsibilities and accountability — for the accuracy, integrity and originality of the work — that can only be attributed to and performed by a human. Any generative AI use has to be disclosed in the declaration statement above, but the tool itself is never listed in the byline or acknowledged as an author. This is consistent with the position taken by the Committee on Publication Ethics (COPE) and by the International Committee of Medical Journal Editors (ICMJE), both of which hold that AI tools cannot meet authorship criteria because they cannot take responsibility for a work’s content.
Figures, images and graphical abstracts
Elsevier draws a firm line between AI-assisted visualization of real data and AI-generated imagery standing in for actual research output.
| Generally permitted (with disclosure) | Not permitted |
|---|---|
| Explanatory images such as flowcharts, diagrams and decision trees — disclose in the figure caption and in the AI declaration | Primary research images: microscopy, western blots, radiology/imaging scans, patient photographs |
| Data visualizations generated directly from the authors’ own underlying data, described in the Methods section | Graphical abstracts created with general-purpose generative AI tools |
| AI-assisted research imaging methods used for data collection, described in Methods | Cover art generated with AI, without prior permission from the editor/publisher |
The underlying reasoning is one of scientific integrity: figures that stand in for real research results (a blot, a scan, a micrograph) must show the actual data, not an AI-generated approximation of it, because a reader or reviewer relies on that image as primary evidence.
Peer reviewers: confidentiality comes first
Elsevier’s policy also restricts how peer reviewers can use generative AI. Reviewers must not upload a manuscript, or any part of it, to a generative AI tool — doing so is treated as a breach of the confidentiality of the peer-review process and of the authors’ intellectual property, since most consumer AI tools retain or process submitted text on external servers outside the journal’s control. Reviewers may still use AI tools in a more limited way, such as improving the language of their own review report or running a background literature search, but human judgment and oversight of the actual review has to remain with the reviewer.
Editorial decision-making
Elsevier’s broader position (consistent with the COPE and STM Association guidance referenced above) is that editorial decisions on manuscripts remain a human responsibility. Generative AI is not treated as a substitute for editor or reviewer judgment in accept/reject decisions.
How Elsevier’s policy compares to other major publishers
Elsevier’s core disclosure-and-declaration approach — a required, published statement, no AI authorship, an exemption for basic language checks — is broadly consistent with the direction most large publishers have taken, though the exact wording, placement and figure rules differ. See the equivalent pages for SAGE’s generative-AI and authorship policy, Taylor & Francis’s generative-AI authorship policy, and MDPI’s generative-AI authorship policy for the same breakdown at those publishers, and the IEEE Generative AI Policy and ICMJE generative AI recommendations dictionary entries for the standards-body positions Elsevier’s policy aligns with.
Frequently asked questions
Does Elsevier require AI disclosure?
Yes. Any use of generative AI or AI-assisted technologies in preparing the text of a manuscript must be disclosed in a dedicated declaration statement placed before the references, using Elsevier’s suggested wording or an equivalent statement covering the same information (tool used, purpose, and author review/responsibility).
Can I use ChatGPT to write my Elsevier paper?
Elsevier’s policy does not prohibit using a generative AI tool to help draft or improve manuscript text, but it must be disclosed in the declaration statement, and the author(s) must review the output and take full responsibility for the accuracy and integrity of the final content. The AI tool cannot be listed as an author.
Where do I put the AI declaration in an Elsevier manuscript?
At the end of the manuscript, immediately before the reference list, under the heading “Declaration of generative AI and AI-assisted technologies in the manuscript preparation process.”
Do I need to disclose Grammarly or basic spell-checking?
No. Elsevier’s policy explicitly exempts basic checks of grammar, spelling and punctuation from the disclosure requirement.
Can AI be listed as a co-author on an Elsevier journal?
No. Elsevier’s policy states that authorship carries responsibilities and accountability that can only be attributed to and performed by humans, so a generative AI tool cannot be listed as an author or co-author under any circumstances.
Can peer reviewers use AI tools to review a manuscript for Elsevier?
Reviewers must not upload manuscripts to generative AI tools, since doing so breaches the confidentiality of the review process and the authors’ intellectual property rights. Limited use — such as improving the language of a reviewer’s own report or a background literature search — is permitted, provided the reviewer’s own judgment governs the actual review.
Last verified 2026-08-16 against Elsevier’s own generative-AI policy page (elsevier.com/about/policies-and-standards/generative-ai-policies-for-journals). Publisher AI policies are an actively evolving area — check the live policy page for the current version before submission if this page is more than a few months old.
