Generative AI Policies for Journals

Generative artificial intelligence and AI-assisted technologies (collectively, “AI tools”) are increasingly used in research and scholarly publishing. These technologies may offer useful support, but their use must be responsible, transparent, and consistent with research integrity, confidentiality, intellectual property, and data protection requirements.

This policy explains the standards that apply to authors, reviewers, editors, and Tech Science Press (“TSP”). TSP will review and revise this policy as relevant technologies, professional standards, and publishing practices evolve.

1. Authors

1.1 Use of AI tools in manuscript preparation

TSP recognizes that, when used appropriately, AI tools can assist researchers with tasks such as surveying the literature, exploring research questions, identifying knowledge gaps, organizing material, and improving language or readability. Such assistance must not replace the authors’ own expertise, critical analysis, interpretation, or judgement.

Authors must retain meaningful human oversight throughout manuscript preparation and remain fully responsible for the submitted and published work. In particular, authors must:

· verify the accuracy, completeness, and balance of AI-assisted content, including all citations and references, which may be inaccurate or fabricated;

· revise AI-assisted material as necessary so that the manuscript reflects the authors’ original contribution, reasoning, and conclusions;

· disclose relevant use of AI tools in accordance with Section 1.3; and

· protect confidential information, personal data, intellectual property, and third-party rights by reviewing the applicable terms and privacy practices of each tool.

1.2 Responsible use

Before using an AI tool, authors must determine how the provider collects, retains, reuses, or shares submitted content. Unpublished manuscripts, confidential material, and personally identifiable information should not be entered into a tool unless adequate privacy and confidentiality safeguards are in place.

Authors should also confirm that the tool’s terms do not claim any rights over the submitted material or generated output that could conflict with the authors’ rights, third-party rights, or the subsequent publication of the article. Any factual error, misleading statement, bias, or inappropriate output introduced through an AI tool must be identified and corrected by the authors.

1.3 Disclosure and authorship

Any material use of AI tools must be disclosed in the manuscript at the time of submission, as appropriate, in the Methods section and/or the Acknowledgement section. The disclosure should identify the tool used, explain the purpose for which it was used, and describe the authors’ review and oversight.

The following principles apply:

· Routine spelling, grammar, or punctuation checks do not require disclosure.

· AI-assisted changes that substantially affect the wording, sentence structure, organization, or other content of the manuscript must be disclosed in the Acknowledgement section.

· Specialist assistive technologies used solely for accessibility purposes do not require disclosure.

· When an AI tool forms part of the research design or methods, including data collection, analysis, or interpretation, its use must be described in sufficient detail in the Methods section in accordance with Section 1.5. Suggested acknowledgement statement:

Acknowledgement: During the preparation of this manuscript, the author(s) used [NAME OF TOOL/SERVICE] for [PURPOSE]. The author(s) subsequently reviewed and revised the relevant content and accept(s) full responsibility for the article.

AI tools cannot be named as authors or co-authors and must not be cited as authors. Authorship requires human accountability, approval of the final manuscript, agreement to submission, and the ability to address questions about the accuracy and integrity of the work. Authors must also ensure that the work is original, that every named author meets the applicable authorship criteria, and that the manuscript does not infringe third-party rights. Authors should consult TSP’s publishing ethics policies before submission.

1.4 AI tools in figures, images, and artwork

Whether AI tools may be used for visual material depends on the purpose and evidentiary role of the image.

1.4.1 Explanatory images

Explanatory images communicate concepts, processes, relationships, or workflows—for example, flow charts, decision trees, timelines, conceptual schematics, and experimental workflow diagrams. AI tools may assist in preparing these images, provided that the requirements for human oversight, accuracy, transparency, disclosure, and authorship in this policy are met.

Each affected image caption should identify the tool and version. The use should also be disclosed in the manuscript’s Acknowledgement section.

1.4.2 Research and data images

Visual material that reports research data or findings includes both data visualizations produced from datasets and primary images that record observations or experiments.

1.4.2.1 Data visualizations

Plots, charts, graphs, heatmaps, and similar visualizations must faithfully represent the underlying data. AI tools must never be used to invent or alter data, fabricate findings, or create a visual result that cannot be supported by the reported data and methods.

AI-assisted data visualizations are permitted only when they are generated directly from the underlying data through reproducible analytical, computational, or statistical procedures. The Methods section should describe those procedures and, where applicable, state the name and version of the tool or model and its provider or developer. Authors must follow relevant licence and attribution requirements. TSP may request records of the AI-assisted workflow, the underlying data, or original unprocessed images for editorial review.

1.4.2.2 Primary research images

Images such as microscopy, histology, western blots, radiological scans, and patient images constitute primary research evidence. AI tools must not be used to generate observations that were not obtained in the research or to alter the content of such images, including through AI-based changes to brightness, contrast, or colour balance.

1.4.3 Graphical abstracts and cover artwork

Graphical abstracts and cover artwork summarize or visually represent an article but do not constitute primary scientific evidence.

General-purpose generative image tools may not be used to produce graphical abstracts. Authors are encouraged to use scientific illustration tools with clear licensing arrangements and to ensure that every visual element is accurate and appropriately licensed.

AI-generated or AI-assisted journal cover artwork may be considered only with advance approval from the journal editor and TSP. Authors should demonstrate that they have the rights and permissions required for all source material and should provide appropriate attribution.

1.4.4 Responsibilities for visual material

Authors are accountable for the accuracy, integrity, and originality of all submitted visual material. If an image incorporates or is derived from existing artwork or graphics, the source must be acknowledged and any necessary permission obtained before submission. Authors must not use AI tools to imitate protected images, identifiable individuals, voices, products, or brands in a manner that infringes rights or misleads readers.

The requirements in Section 1.2 also apply to all AI-assisted visual material.

1.5 AI tools used as research methods

The restrictions in Section 1.4 do not prohibit the legitimate use of AI tools as part of the research design or methods. Examples include AI-assisted data collection, image acquisition or interpretation, data analysis, and predictive modelling.

Such use must be described in sufficient detail in the Methods section to allow readers to understand and evaluate the role of the AI tool and, where feasible, reproduce the research. Authors should identify the tool or model, its version, and its provider or developer, as applicable, and explain how it was used. They must also comply with the tool’s applicable terms and provide appropriate attribution where required.

Authors remain responsible for verifying the validity, accuracy, and reliability of AI-assisted methods and outputs. AI tools must not be used to fabricate, manipulate, or misrepresent research data or findings.

2. Reviewers

2.1 AI use during peer review

Peer review depends on confidentiality, independent expert judgement, and accountability. Manuscripts and all related submission materials provided for review must be treated as confidential.

Reviewers must not independently upload a manuscript, any part of it, or related materials—including figures, tables, and supplementary files—to an AI tool. An exception applies only where TSP has expressly authorized the tool and appropriate confidentiality, privacy, and security safeguards are in place.

The scientific assessment, comments, and recommendations in a review report must remain the reviewer’s own. AI-generated findings may be inaccurate, incomplete, or biased and must not replace the reviewer’s critical analysis or determine the reviewer’s recommendation. Reviewers remain fully responsible for the content and conclusions of their reports.

Subject to the confidentiality requirements above, reviewers may use an AI tool to improve the clarity or organization of text they have written. Reviewers must maintain human oversight and verify all AI-assisted output.

3. Editors

3.1 AI use in editorial work

Editorial assessment and publication decisions require human responsibility and judgement.

Editors should not upload a manuscript or any confidential part of a submission to an AI tool unless TSP has expressly approved that tool and the required confidentiality, privacy, and security safeguards are in place.

AI tools must not replace an editor’s independent evaluation or determine the final editorial decision. Editors remain accountable for the assessment process, the decision, and its communication to the authors.

Subject to the confidentiality requirements above, editors may use an AI tool to improve the clarity or organization of text they have written. Editors must maintain human oversight and verify all AI-assisted output.

4. Use of AI by Tech Science Press

TSP may use AI-assisted systems to support selected publishing activities, which may include:

· suggesting potentially suitable reviewers;

· assessing whether a submission aligns with a journal’s scope;

· identifying possible duplicate submissions;

· performing technical and completeness checks against submission requirements;

· supporting research-integrity and policy-compliance screening; and

· assisting post-acceptance production, including copyediting, proof preparation, and the identification of possible inconsistencies or errors.

AI-assisted results provide supporting information and do not constitute independent editorial or publication decisions. Human oversight remains central, and responsible personnel retain authority and accountability for all decisions. TSP evaluates relevant tools for reliability and bias and requires their use to comply with applicable confidentiality, data-protection, and information-security obligations.

The type and extent of AI-assisted support may differ among TSP journals. Questions about the practices of a particular journal should be directed to that journal’s editorial office.