Artificial Intelligence

Artificial Intelligence in Academic Publishing

Artificial intelligence (AI) tools, including generative AI and large language models (LLMs), can support a wide range of activities related to research and academic publishing. They can be used, for example, for planning work, searching for and organizing information, working with text, programming, data analysis, or creating visual outputs.

The Guidelines for the Use of Artificial Intelligence at Mendel University in Brno therefore do not apply solely to text generation. They cover textual outputs, including scholarly articles, source code and scripts, data analyses and computational outputs, graphics, photographs, video, audio, animations, and other materials created in connection with activities at MENDELU.

MENDELU supports the meaningful, critical, and ethical use of AI as a modern tool. However, the use of AI must comply with the principles of academic integrity, data protection, copyright and licensing requirements, and must remain under human responsibility.

In the context of academic publishing, it is also important to note that MENDELU’s rules do not replace the publication policies of individual journals or publishers. Before submitting a manuscript, it is therefore always necessary to check the current author guidelines of the relevant journal.

The goal of Mendel University is the rational, conscious and ethical use of artificial intelligence by students and staff. To maintain high academic and ethical standards, the use of AI is regulated in these documents:

MENDELU Principles for the Use of AI

When using AI in research, manuscript preparation, or the creation of other academic outputs, the following principles apply at MENDELU in particular:

AI should support the author’s work, not replace the author. Uncritical adoption of outputs that the author did not create themselves is inconsistent with academic standards. AI must not replace the author’s own creative work, scholarly argumentation, interpretation of results, or other key parts of the work.
The use of AI must be transparent. If AI was used in the creation of an output, its use must be appropriately disclosed. Undisclosed use of AI constitutes a breach of academic integrity and may be considered plagiarism.
The author always bears responsibility for the final content. Responsibility cannot be transferred to an AI tool or its provider. The author is also responsible for any parts of the output that were created, reformulated, or otherwise influenced by AI.
AI-generated outputs must be critically verified and revised. The author must check factual accuracy, argumentation, sources used, citations, analytical results, and any other information provided by AI. AI may generate inaccurate or entirely fabricated information and references.
Copyright and licensing rights must be respected. The use of AI does not remove the obligation to properly cite original sources, comply with licences, and respect copyright, database rights, and other third-party rights.

These principles are based on the Guidelines for the Use of Artificial Intelligence at Mendel University in Brno from 2026.

Data Security When Using AI

Before entering any information into an AI tool, it is necessary to assess the nature of the information and the terms and conditions of the specific service.

When working with AI, applicable laws and regulations, particularly the GDPR, MENDELU rules, and the rules governing the handling of individual categories of data, must be observed.

Do not enter personal or sensitive data, contractual documents, internal MENDELU documents, or other protected data into publicly available AI tools that do not provide appropriate protection of user data.

According to MENDELU’s Guidelines, non-public research data and internal MENDELU documents that do not contain sensitive data may be used in internal AI tools. However, the specific data category, the licensing terms of the tool being used, and any third-party rights must always be respected.

More information is available on the Open Science Centre’s Data Categorisation page. Information about AI tools available under university licences is provided by OIT.

Special Considerations for Unpublished Manuscripts

Similar caution should also be exercised with manuscripts, peer-review reports, grant proposals, and other non-public materials. Before entering such content into an AI tool, you must verify that you have the right to provide it to the service and understand how the provider handles the submitted data.

Many academic publishers explicitly prohibit reviewers or editors from uploading non-public manuscripts to publicly available generative AI tools in order to protect confidentiality, copyright, and personal data.

Use of AI in Academic Manuscript Preparation

Policies vary across publishers, but current publication guidelines commonly include the following principles:

AI tools generally cannot be listed as authors or co-authors because they cannot assume responsibility for published content;
AI may be used as a supporting tool, but it should not replace the author’s expert judgment, interpretation of results, or formulation of scientific conclusions;
many publishers require the use of generative AI to be transparently disclosed;
publishers may require authors to specify the name and version of the tool, the purpose for which it was used, the parts of the manuscript affected by AI, or how the AI-generated output was checked;
substantially stricter rules may apply to the generation or modification of images, data, or other research outputs than to language editing of text.

Specific requirements may vary not only between publishers but also between individual journals published by the same publisher. Always check the current Instructions for Authors / Editorial Policies of the specific journal immediately before submitting your manuscript.

Overview of Policies of Selected Academic Publishers

Information verified as of 14 August 2026. This overview is for guidance only and does not replace the current policies of the individual journal.

Publisher	Key rules for the use of AI
American Chemical Society (ACS)	AI cannot be listed as an author. The use of AI to generate text or images should be disclosed in the Acknowledgments; more extensive use should also be described in the Methods or another appropriate section. For AI-generated graphics, ACS also requires transparency in the figure caption.
Cambridge University Publishing & Assessment	The use of AI should be transparently disclosed and explained. AI does not meet the requirements for authorship, and authors remain responsible for the accuracy, integrity, and originality of their work. The use of AI must not violate policies on plagiarism.
DeGruyter	AI cannot be listed as an author, and the publisher does not accept works whose content has been primarily generated by AI. Significant use of AI in manuscript preparation must be disclosed, and the author remains fully responsible for the entire output.
Elsevier	Generative AI may support manuscript preparation but must not replace the author’s expert judgment. AI cannot be listed as an author. The use of generative AI in manuscript preparation must be disclosed in accordance with the publisher’s guidelines, and authors must verify and revise the resulting output.
Inderscience	AI cannot be listed as an author. If AI has been used to create, write, generate, or edit article content, the tool used and its purpose must be described in the Acknowledgements. The publisher exempts routine spelling and grammar checking tools and reference managers from this disclosure requirement.
MDPI	Generative AI and LLMs cannot be listed as authors. Authors are responsible for the originality, accuracy, and integrity of the manuscript. The use of GenAI must be disclosed, including whether it was used, for example, for text, data, graphics, study design, or data analysis and interpretation.
MIT Press	AI tools cannot be listed as authors. Authors who use AI to create text, images, or graphics, or in data collection, should inform editors and transparently describe its use in the manuscript.
Oxford University Press	AI cannot be listed as an author. For journal manuscripts, the use of AI should be disclosed to editors and, depending on the type of use, described in the Methods or Acknowledgements. OUP also provides separate, more detailed guidance for books and other publication formats
PLOS	Any contribution of AI to the study or article content must be clearly described in the Methods or, where appropriate, in the Acknowledgements. Authors should specify the tool used, how it was used, how the output was verified, and which parts of the study or manuscript were affected by AI.
Sage Publications	Sage distinguishes between assistive and generative uses of AI. Routine assistance with language, grammar, or structure may not require disclosure under the general policy, whereas AI-generated content must be disclosed. Individual journals may impose additional requirements.
Science (AAAS)	Science journals apply relatively strict rules. AI cannot be listed as an author; the use of AI in manuscript preparation must be transparent, and specific restrictions apply to AI-generated images and multimedia. Authors should always check the current policy of the relevant Science journal before submission.
Springer Nature	Springer Nature supports the responsible and transparent use of AI, but responsibility for scholarly content and decision-making must always remain with humans. AI may support the work but must not replace expert judgment. Additional restrictions apply to AI-generated visual materials.
Taylor & Francis	AI may be used in research and manuscript preparation under specified conditions. Where AI is used as part of the methodology, sufficient details and appropriate disclosure must be provided. Authors must verify the accuracy of AI outputs and retain records of their process if requested by the publisher. AI must not replace the fundamental responsibilities of the researcher and author.
The Royal Society	The use of AI must be disclosed in the manuscript, and the relevant statement may be published alongside the article. AI cannot be listed as an author and must not replace key research tasks such as expert analysis, interpretation, or the formulation of conclusions.
Wiley	AI may be used as a supporting tool, but not as a substitute for the author’s expertise and judgment. Authors remain fully responsible for the work, and significant use of AI in the creation, revision, or translation of a manuscript must be transparently disclosed. Wiley has published more detailed and updated guidance for researchers.

Publisher policies are evolving rapidly. Therefore, the current policy of the specific journal at the time of manuscript submission should always be regarded as authoritative, rather than relying solely on this overview.

How to Disclose the Use of AI in a Manuscript

The form and placement of an AI disclosure are determined by the specific publisher or journal. The disclosure may, for example, take the form of a separate statement, a statement in the Methods or Acknowledgements section, or information provided in the cover letter or in a form completed during manuscript submission.

If the publisher provides its own wording or template, its instructions should be followed as a priority.

In general, an AI disclosure should include:

the name of the AI tool used and, where relevant, its model or version;
the purpose for which the tool was used;
the stage of the research or the parts of the manuscript affected by the use of AI;
how the author verified and revised the AI-generated outputs;
any additional information required by the publisher, such as the date of use, prompts, or parameters used in the analysis.

General Example of a Disclosure Statement

During the preparation of this manuscript, the authors used [tool and version/model] for [purpose] during [stage/section of the work]. All AI-assisted outputs were critically reviewed and edited by the authors, and factual claims, sources, and references were independently verified. The authors take full responsibility for the final content of the manuscript.

The specific wording should always be adapted to the publisher’s requirements and to the actual way in which AI was used.

If AI Was Used as Part of the Research Methodology

If AI is not merely used as an aid in writing but forms part of the research methodology itself—for example, in data analysis, classification, data generation or processing, programming, image-data processing, or another reproducible procedure—its use should be described in the same way as any other research tool or software.

The description should enable the reader to understand what tool was used, what inputs were provided, for what purpose it was used, what outputs it produced, and how the results were verified.

AI, Citations, and References

An AI tool is not a substitute for a scholarly information source.

If an AI tool recommends a publication, DOI, author, statistical figure, or other claim, the author must verify the information in the original source. The reference list should contain the actual sources on which the claims are based, rather than unverified AI output.

The MENDELU Guidelines explicitly maintain the author’s obligation to cite sources and comply with licensing requirements even when AI is used during the work.

Use of AI in Peer Review

Specific rules apply to peer review.

A manuscript submitted for peer review is generally a non-public and confidential document. Reviewers should therefore not upload the manuscript or any part of it to a publicly available generative AI tool unless this is explicitly permitted by the publisher’s policies and all applicable data-protection and confidentiality requirements are met.

Some publishers explicitly prohibit the use of publicly available generative AI in peer review, while others permit only limited use under specified conditions and subject to disclosure. It is therefore essential to follow the policies of the relevant publisher. Cambridge, for example, states that no part of a manuscript should be shared with a publicly available generative AI tool without specific permission; similar restrictions are imposed by other publishers.

Detection of AI-Generated Content

Tools that attempt to determine whether a text was generated using generative AI have limited reliability.

According to the MENDELU Guidelines, AI-detection tools may be used only as a supporting indicator. On their own, they cannot be regarded as evidence of a breach of academic integrity.

The output of an AI detector should therefore not be interpreted as conclusive evidence of authorship or unethical use of AI.

It is also important to bear in mind that uploading text to an external detection tool may involve providing content to a third party. In the case of non-public manuscripts, internal documents, personal data, or other protected content, the service’s terms and data-protection arrangements should therefore be assessed in advance.

Last updated: 14 August 2026

We use cookies

Artificial Intelligence