What Journals Now Expect When Authors Use AI, and How to Check Your Own Manuscript
The First Art Newspaper on the Net    Established in 1996 Wednesday, November 18, 2026


What Journals Now Expect When Authors Use AI, and How to Check Your Own Manuscript





A postdoc is three days from submitting her first paper as lead author. She used a language model to tighten two paragraphs of the introduction, and she isn't sure whether that needs mentioning. The journal's instructions run to nine pages. She finds the AI paragraph on page seven.



Policies like that one are now common across scholarly publishing, and the rules are clearer than most researchers realise. This article walks through what a major publisher says about authorship, disclosure and peer review, and offers a practical way to audit your own manuscript before you submit.



Language models cannot be authors



The Springer Nature policy, published on the npj Quantum Information editorial pages, is direct. It states that large language models, such as ChatGPT, "do not currently satisfy our authorship criteria," because authorship carries accountability that "cannot be effectively applied to LLMs" (Springer Nature, as published on npj Quantum Information).



The logic is simple. An author takes responsibility for the accuracy, integrity and originality of a work, answers readers' questions and corrects errors. A language model can't do any of that. So a model never appears in an author list, however much it contributed to a draft.



Document your use in the Methods



The policy goes on to say that use of an LLM "should be properly documented in the Methods section," or in a suitable alternative section if the manuscript has none. For work involving non-generative machine learning tools that manipulate or enhance images, the policy asks for disclosure in the relevant caption.



In practice this means a short, factual statement. Name the tool and version, say what you used it for and confirm that the authors reviewed and take responsibility for the result. If you used it to summarise sources, draft a section or generate code, say so. A sentence or two is typically enough, and it spares you an awkward exchange with an editor later.



Remember that the rules differ between publishers and between journals. The postdoc's nine pages might say something slightly different from the example above, so the author guidelines of your target journal are the final authority.



Reviewers have obligations too



Authors aren't alone in being addressed. The same policy asks that peer reviewers do not upload manuscripts into generative AI tools while the publisher explores safe options. If AI supported any part of a reviewer's evaluation of a manuscript's claims, the policy asks reviewers to declare that use transparently in the review report.



The reasoning is confidentiality. An unpublished manuscript belongs to its authors, and pasting it into an outside tool can expose it. For researchers who also review, the same principle applies to every file marked confidential.



A short self-audit before you submit



A manuscript can pass editorial checks and still carry loose ends about AI. Before you submit, spend an hour on this routine.




  • Reread the author guidelines for AI wording, then compare them with what you did.

  • List every AI-assisted step and where it appears in the manuscript.

  • Verify every claim and citation that passed through a model, since fabricated references are a known risk.

  • Draft the disclosure sentence and place it where the journal expects it.

  • Check the prose for flat, formulaic passages and revise them in your own words.



It's reasonable to check for AI patterns in your abstract and introduction at this stage, as one more read before submission. Phrasly's page claims its tool can distinguish text written by AI from text a person wrote and then edited with AI tools. That is a vendor claim, so treat it with the usual caution, and don't treat a result as a substitute for a Methods statement. A score can't tell an editor what you did. Your disclosure can.



What to do about collaborators



Multi-author papers raise a quieter issue. One co-author may have used a tool without telling the others, and every name on the paper carries responsibility. Agree on a shared rule early, ask each author to note any AI use in a shared log and review the log before submission. It takes five minutes, and it prevents a retraction request from landing on a colleague's desk.



Where the rules still vary



Publishers don't agree on every detail. Check whether your target journal treats light language editing differently from content generation, where the disclosure statement should go and whether figures are handled under a separate rule. If the instructions are unclear, email the editorial office. Editors would rather answer a question in week one than untangle a problem in week twelve, and a short, polite query shows you're taking the rules seriously.



A sample disclosure



A disclosure can be as short as this: "The authors used a large language model to improve the clarity of two paragraphs in the Introduction. All suggestions were reviewed and edited by the authors, who take full responsibility for the content." Adapt the wording to what you actually did, and never copy a statement that describes something you didn't do. Accuracy matters more than polish here.



Why early-career researchers should care most



Senior authors have track records that cushion an awkward query. Early-career researchers don't. A paper flagged for undisclosed tool use, or for a citation a model invented, can follow a junior name for years. Getting the disclosure right at the start costs a couple of sentences, and getting it wrong can cost a reputation that's still being built.



A note on figures and code



The same questions apply to figures, code and data analysis. If a model helped write a script, say so, and make sure someone on the team has run and understood it. Reproducibility depends on knowing exactly what produced a result, and an undisclosed tool in the chain makes that harder to audit.



Disclose First, Detect Second



The direction of travel in publishing is plain: models aren't authors, use is documented in the Methods and confidential manuscripts stay out of public tools. A detector can help you notice weak or generic passages in your own writing, but it can't replace honest disclosure. The postdoc added two sentences to her Methods, fixed a paragraph that sounded oddly generic and submitted on schedule.



If you'd like more help with the practical side of writing for publication, further reading on the Phrasly blog offers guides on drafting, revising and checking academic work.



Today's News

November 12, 2026

Charlotte Jackson Fine Art presents Helen Pashgian

Opera Gallery Miami pairs Alex Katz and Julian Opie in exhibition

Gagosian presents Georg Baselitz's final paintings and sculptures in New York

The Frick Collection presents Kent Monkman paintings inspired by Vermeer

Hauser & Wirth announces representation of artist rafa esparza

Roger Dean's original 'Fragile' cover art for Yes emerges after decades missing to make auction debut at Heritage

Tate Britain celebrates creative partnership of Vanessa Bell and Duncan Grant

Albertina presents major retrospective of Austrian artist Arnulf Rainer

Prada presents 'Chuanqi: Spacetime Dialogues' exhibition at Shanghai's Rong Zhai

Sotheby's Important Watches sale breaks new records in Geneva

Sarah Morris examines corporate power in new São Paulo exhibition

Galerie Eva Presenhuber presents Louisa Gagliardi's fifth solo exhibition

Kunsten Museum of Modern Art Aalborg exhibits six decades of Danish art and design history

Eunnam Hong explores multiplicity of the self in cinematic paintings at Mendes Wood DM

Heritage Auctions realizes $2.58 million led by record David Bates painting

Alison Bradley Projects presents first exhibition with textile artist Hiroko Takeda

Rago Wright announces special auction celebrating Lindsey Adelman Studio's 20th anniversary

Rare Ionia electrum stater leads Heritage's World & Ancient Coins Auction

Gooding Christie's sells first Ferrari 499P Modificata at auction

Christie's London presents 'Love & Lust' auction of romantic artifacts

Acquavella Galleries presents Harumi Klossowska de Rola's New York solo debut

Abbas Akhavan's first mid-career survey and first major U.S. show opens at the Walker Art Center

Museum Folkwang explores the interconnections between Islamic art and European modernism

Museum Angewandte Kunst explores Japanese aesthetic of shibui in new exhibition

What Journals Now Expect When Authors Use AI, and How to Check Your Own Manuscript

Offlico Is the Best Appointment Reminder Software for Field Service Teams and Sole Traders

Global Satellite Backhaul Market to Reach USD 9.20 Billion by 2033, Driven by Rising Demand for Maritime Connectivity




Museums, Exhibits, Artists, Milestones, Digital Art, Architecture, Photography,
Photographers, Special Photos, Special Reports, Featured Stories, Auctions, Art Fairs,
Anecdotes, Art Quiz, Education, Mythology, 3D Images, Last Week, .

 



The OnlineCasinosSpelen editors have years of experience with everything related to online gambling providers and reliable online casinos Nederland. If you have any questions about casino bonuses and, please contact the team directly.


Abogado de Accidentes



Founder:
Ignacio Villarreal
(1941 - 2019)


Editor: Ofelia Zurbia Betancourt

Art Director: Juan José Sepúlveda Ramírez


Royalville Communications, Inc
produces:

ignaciovillarreal facundocabral-elfinal.org
Founder's Site. Hommage
       
Tell a Friend
Dear User, please complete the form below in order to recommend the Artdaily newsletter to someone you know.
Please complete all fields marked *.
Sending Mail
Sending Successful