AI Use & Disclosure
Applies to the ESS Press book line: books, monographs, chapters, and conference proceedings. Journal articles follow the rules of the respective journals on their websites. The purpose of this policy is to ensure research integrity, transparency, and lawful use of AI tools throughout all stages of content preparation and publication.
1) Principles and scope
1.1. ESS Press supports the responsible use of AI tools (including LLMs/generative models, machine translation, and image/data analysis tools), provided that use is transparent, outcomes are controlled, and rights are respected.
1.2. This policy is addressed to authors, editors, guest editors/volume editors, and reviewers, and it also describes permissible AI-related procedures within the publisher’s production cycle.
1.3. Where this Policy conflicts with an agreement and/or funder requirements, the stricter rules prevail. Journal publications are governed by the journals’ separate websites.
2) General requirements for AI use
2.1. AI is not an author. AI systems are not recognized as authors and cannot be listed as co-authors. Responsibility for the content rests entirely with the named human authors.
2.2. Prohibition of falsification. It is not permissible to use AI to generate, substitute, or “invent” data, references, quantitative results, or images/figures as if they were empirical outputs, unless clearly labeled as AI-generated and methodologically justified.
2.3. Transparency. Any substantive use of AI in drafting, translation, analysis, or image processing must be disclosed.
2.4. Verification. Authors must verify facts, calculations, quotations, and references produced with AI assistance. Hallucinations and erroneous attributions are not acceptable.
2.5. Rights and licensing. AI-generated content may be included only if copyright, licensing, and ethical constraints are respected (see §6 and the “Author Rights and Licensing” policy).
3) Permitted uses (with disclosure)
3.1. Language editing and style: grammar, clarity improvements, terminology harmonization.
3.2. Machine translation: as a draft stage, followed by specialist review and editing.
3.3. Technical assistance: drafting preliminary keyword lists, section structure suggestions, summaries of the author’s own materials, letter templates.
3.4. Auxiliary image/data processing: format conversion, metadata extraction, preliminary segmentation—provided that a human verifies outputs and originals are preserved.
3.5. Reference/metadata reformatting: aligning citations with a required style, checking DOIs (with mandatory manual validation).
4) Prohibited or restricted practices
4.1. Generating “data” (numerical series, images, micrographs, maps, etc.) presented as empirical evidence without explicit labeling and a methodological description is prohibited.
4.2. Fabricated references/quotations (“reference hallucinations”) are prohibited.
4.3. Undisclosed manipulation of images/figures (removing/adding objects, changing scale/intensity) is prohibited.
4.4. Unauthorized uploading of confidential materials (manuscripts, datasets containing personal information) to external AI services is prohibited unless explicit permission and appropriate safeguards/anonymization are in place.
4.5. Automated drafting of literature review sections without critical verification of primary sources is not permitted.
4.6. Using AI to circumvent peer review (e.g., generating “responses to reviewers” without substantive methodological work) is not permitted.
5) Requirements for reviewers and editors
5.1. Confidentiality. Manuscripts and related materials must not be uploaded to external AI services (including cloud-based LLMs) unless the publisher/project has contractually secured safe solutions. Limited local grammar assistance for a reviewer’s own report is acceptable, provided that manuscript text is not transmitted.
5.2. Transparency. If AI was used in preparing a review (e.g., to improve the language of the report), the reviewer should note this in a private comment to the editor.
5.3. Responsibility. Editors must not make decisions based solely on automated reports. Screening outputs (including AI tools for text/images) require human interpretation.
6) Rights, licensing, and AI-generated content
6.1. Authorship and attribution. AI tools may be acknowledged in “Methods” and/or “Acknowledgements,” but not in the author list.
6.2. Third-party materials. If an AI model may reproduce styles or elements protected by copyright (images/text), the author is responsible for lawful inclusion and license compatibility.
6.3. Licensing of outputs. AI-generated text/images incorporated into a Work are licensed under the author’s chosen license (e.g., CC BY 4.0 for OA). The author warrants that no third-party rights are infringed.
7) Disclosure of AI use (Author Disclosure)
7.1. Where AI was used substantively (beyond simple spelling corrections), authors must include a disclosure statement in the manuscript (at the end of the chapter/book or in a methodological appendix) and in submission documents.
7.2. The disclosure should specify purpose, tools, versions, extent of use, and the nature of human oversight.
7.3. For AI-assisted translation, authors must indicate the source language, the tool, and the fact of subsequent human review/editing.
8) Data, humans/animals, and AI
8.1. AI processing of personal data is permitted only where lawful bases exist, ethics approval is obtained where applicable (IRB), and de-identification safeguards are implemented. This must be described in the Data Availability Statement and/or Methods.
8.2. For animal-related images/data, relevant standards apply; any transformations must preserve scientific meaning and be reproducible.
9) ESS Press production processes
9.1. The publisher may use local and/or contractually secured AI tools for technical production support (e.g., reference checks, typography, layout validation, metadata matching), without altering scholarly content.
9.2. Any edits affecting meaning are agreed with the authors during proofing.
9.3. Screening checks (text similarity, image manipulation detection) are used as editorial support and are not an end in themselves; final decisions are made by a human editor.
10) Non-compliance and measures
10.1. Undisclosed or improper AI use is handled under the Publication Ethics Policy. Measures range from requests for correction/disclosure to rejection or retraction where the scholarly record is materially distorted.
10.2. Where fabricated data/references, unethical image manipulation, or confidentiality breaches are identified, an investigation is initiated under COPE-aligned procedures.
11) Disclosure statement templates (to insert into the manuscript)
11.1. Language editing (EN):
“Language editing assistance was provided by AI tools [name, version] to improve clarity and grammar. All facts, data, and references were verified by the authors; the authors take full responsibility for the content.”
11.2. Machine translation (EN):
“A preliminary translation was generated using [tool, version] and subsequently fully reviewed and edited by the authors/editor.”
11.3. Image processing (EN):
“[Tool, version] was used for contrast normalization/segmentation; adjustments did not alter the scientific meaning. Original images are available upon request/in the repository (see DAS).”
12) Confidentiality and security
12.1. Authors/reviewers/editors must assess data-leakage risks when using cloud AI services; confidential materials must not be uploaded without explicit permission and appropriate safeguards.
12.2. A local or contractually secured environment is recommended; access logs and version control are part of internal quality practice.
13) Relationship to other documents
This AI Policy complements: Publication Ethics, Peer Review Policy, Author Rights and Licensing, Data Availability, Open Access & Self-Archiving, and the Privacy Policy.
14) Dates and contacts
Version: 1.0 · Effective date: [insert publication date of the page]
Questions and consultations: policies@ess.press
Violations/incidents: policies@ess.press