AI Policies: Journals, Funders, and Publishers

Research compiled March 2026. For E1 (scholars mini-book) and PhD section of E2.

The short version

The universal rules across all publishers: 1. AI cannot be listed as an author 2. Disclose if AI materially affected the content 3. Do not use AI in peer review (reviewers: do not upload manuscripts to AI tools) 4. Do not create or manipulate research data, figures, or images with AI 5. You take full responsibility for all AI-assisted content

Grammar and language tools (Grammarly, LanguageTool, DeepL): No disclosure required at most publishers. Exception: OUP requires disclosure even for some assistive tools.

Grant funders

GAČR (Czech Science Foundation)

Status: No AI policy as of March 2026.

GAČR tender documents (2025 and 2026 cycles) contain zero mentions of artificial intelligence, generative AI, or AI disclosure requirements. Researchers preparing GAČR proposals are in a policy vacuum at the funder level.

Practical guidance: Use MUNI/institutional principles as your reference. The EC ERA Living Guidelines (below) are the closest applicable normative standard.

URL: https://gacr.cz/en/extracts-from-tender-documents/

ERC (European Research Council)

Status: Permissive, no mandatory disclosure. Accountability-focused. Date: December 2023 statement; no update found as of March 2026.

• AI use in brainstorming, idea generation, literature search, text revision, translation: acknowledged as normal practice

• No disclosure requirement in applications

• But: "use of external help in preparing a proposal does not relieve the author from full and sole authorship responsibilities"

• Robust text similarity detection in proposals — the ERC can detect AI-generated text

• Practical warning: ERC reviewers report that over-reliance on AI produces generic proposals that score poorly. AI-assisted proposals must still be distinctively yours.

• Evaluators are prohibited from uploading proposals to AI tools

URL: https://erc.europa.eu/news-events/news/current-position-erc-scientific-council-artificial-intelligence

Horizon Europe

Status: Mandatory disclosure in 2025 application forms.

• 2025 proposal forms include an explicit AI declaration field

• Applicants must: declare all AI tools used, provide source lists for AI-generated content, verify plagiarism risks

• "Authors must take full responsibility for all AI-generated content"

• Non-declaration may render a proposal ineligible

• Do not enter patentable or unpublished research into external AI tools without data-sharing disabled

• Evaluators: may not delegate evaluation to AI; AI only for secondary tasks; uploading proposals to AI may violate confidentiality rules

Key practical rule for researchers: Keep a log of AI tool use (when, which tool, how used, settings). This is required for compliance and is good practice anyway.

URL: https://www.emdesk.com/horizon-2020-horizon-europe-basics-guide/horizon-europe-using-ai-in-grant-proposal-writing-responsibly

EC ERA Living Guidelines on Responsible Use of Generative AI in Research

Status: Version 2, April 2025. Regularly updated. Non-binding but the European normative standard.

Core principles for researchers:

• Disclose generative AI use transparently

• Refrain from using generative AI in peer review or grant evaluation

• Respect privacy, confidentiality, intellectual property

• Aim for reproducibility: disclose limitations, biases, hallucination risks of tools used

• Researchers are accountable for AI-generated scientific output

• Comply with EU AI Act

URL: https://research-and-innovation.ec.europa.eu/news/all-research-and-innovation-news/guidelines-responsible-use-generative-ai-research-developed-european-research-area-forum-2024-03-20_en

Czech Academy of Sciences (AVČR)

Status: No AI policy as of March 2026.

The revised Ethics Code (April 2024) does not address AI — the revision focused on ecological footprint, authorship, and harassment. The Commission for Scientific Ethics has acknowledged AI guidelines are needed but has not published them. AVČR researchers are currently guided only by their institution's rules and publisher/funder requirements.

URL: https://www.avcr.cz/export/sites/avcr.cz/.content/galerie-souboru/akademicky-snem/LXIII/Eticky-kodex-vedecko-vyzkumne-cinnosti-v-Akademii-ved-Ceske-republiky.pdf

Publisher policies in detail

De Gruyter / Brill (merged 2024)

• Disclosure required if AI "significantly contributed" to any part of the manuscript or methodology

• Prohibited: AI-generated/manipulated images, figures, or research data; AI in peer review; uploading copyrighted material to AI

• Basic proofreading and language improvement: no disclosure needed

• URL: https://www.degruyterbrill.com/publishing/for-authors/author-policies/artificial-intelligence

Springer Nature / Palgrave Macmillan

• "AI assisted copy editing" (grammar, spelling, formatting): no disclosure

• Any LLM generating text that significantly contributed: disclose in Methods section

• Prohibited: LLM authorship; generative AI images; uploading manuscripts to AI for peer review

• URL: https://www.springernature.com/gp/policies/editorial-policies

Taylor & Francis / Routledge

The dominant humanities/social sciences imprint.

• Disclosure in Methods or Acknowledgments: tool name + version, how used, reason

• Permitted: idea generation, language improvement, literature classification, interactive search

• Prohibited: text/code generation without rigorous human revision; synthetic data substituting real data; creating or manipulating images/figures/data tables

• Individual journals may impose stricter rules — always check the specific journal's instructions

• URL: https://taylorandfrancis.com/our-policies/ai-policy/

Sage

• Grammar/structure tools: no disclosure required

• Generative AI contributing to methodology/results/conclusions: disclose at submission

• Prohibited: AI-generated images; AI in peer review

• Undisclosed use discovered later: may result in rejection at any stage

• URL: https://www.sagepub.com/journals/publication-ethics-policies/artificial-intelligence-policy

Elsevier (includes Social Networks journal)

• Required format for disclosure: a named declaration titled "Declaration of Generative AI and AI-assisted technologies in the writing process," placed before the references

• Required language: "During the preparation of this work the author(s) used [NAME TOOL] 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."

• Prohibited: AI authorship; AI image creation/alteration; copyrighted material in AI without rights; direct AI text without substantial human revision; AI-generated references without verification

• Grammar/spelling only: no disclosure needed

• URL: https://www.elsevier.com/about/policies-and-standards/the-use-of-generative-ai-and-ai-assisted-technologies-in-writing-for-elsevier

Oxford University Press

Notably strictest of the major publishers.

• Substantive generative AI use requires written permission from OUP before submission

• Required disclosure: location in text of each use, tool name and version, manner of use, justification, validation steps

• Prohibited: AI for primary authorial responsibilities (generating arguments, insights, analysis, conclusions) without OUP permission

• Do not upload OUP-intended work to any AI tool that does not permit opting out of training data use

• Digital Scholarship in the Humanities (DSH) is an OUP journal — this policy applies

• URL: https://academic.oup.com/pages/for-authors/books/author-use-of-artificial-intelligence

Cambridge University Press (includes Network Science journal)

• Disclose AI use "just as scholars do with other software, tools and methodologies"

• AI cannot be an author

• First major publisher to issue AI ethics policy (March 2023)

• Opt-in policy: CUP contacts authors before licensing their published content to LLM training datasets

• URL: https://www.cambridge.org/news-and-insights/cambridge-launches-ai-research-ethics-policy

COPE (Committee on Publication Ethics)

The standard most publishers reference. Use this when uncertain about a specific journal.

• AI cannot be listed as author

• Disclose: the full prompt, time/date of query, tool used and its version when AI generated research results

• Grammar/structure tools: no disclosure required

• Verify all AI outputs; check original sources; confirm no hallucinated citations

• Peer review and grant evaluation: refrain from substantial generative AI use

• URL: https://publicationethics.org/guidance/cope-position/authorship-and-ai-tools

Practical checklist for DISSINET researchers

Before submitting a grant proposal:

• GAČR: no funder-specific rules — follow MUNI principles and ERA guidelines

• ERC: no disclosure required, but keep record; ensure proposal is distinctively yours

• Horizon Europe: complete the AI declaration field in the form; keep a log of tools used; verify no unpublished research entered external AI tools

Before submitting a journal article:

• Find your target journal's specific AI policy (check journal website, not just publisher homepage)

• If journal uses T&F/Routledge: check journal-level rules, not just house policy

• If journal uses OUP: contact editorial office before submitting if you used AI substantively

• Prepare your disclosure statement — use the Elsevier format as a safe baseline for any journal that doesn't specify one

• Verify every citation in the manuscript independently — never use Claude-generated references without checking the source exists

• Do not list AI as an author, in acknowledgments as a co-author, or in any way that implies AI authorship

• If you used AI in data analysis or qualitative coding: describe in Methods section (prompts, tools, parameters)

The safe baseline across all contexts:

"During the preparation of this work I used [TOOL NAME, VERSION] for [SPECIFIC PURPOSE]. I reviewed and edited the content produced and take full responsibility for the final text."

Related

• E.muni-guidelines — institutional rules (MUNI)

• E.cuni-researcher-recommendations — CUNI/MUNI joint researcher guidelines (FAIR principles, qualitative disclosure)

• A.critical.limitations — hallucination in citations: why verifying references matters

• A13.examples-dissinet-usecases — data sensitivity for archival and qualitative research