Policy on the Use of Artificial Intelligence

The Ukrainian Journal of Constitutional Law recognises that artificial intelligence (AI), including generative AI and large language models (LLMs), may be used in scholarly research and publishing activities.

The purpose of this Policy is to ensure transparency, academic integrity, reliability of scholarly content, protection of manuscript confidentiality, and a clear allocation of responsibility between humans and technology.

This Policy applies to authors, reviewers, editors and other participants in the editorial and publishing process.

Contents:

1. General Principles

The use of AI tools is not, by itself, a breach of academic integrity. Whether a particular use is acceptable depends on its nature, purpose, extent and transparency.

The Journal follows the following principles:

  • transparency – substantive use of generative AI should be appropriately disclosed;
  • human responsibility – ultimate responsibility for the content, accuracy and integrity of a publication always remains with humans;
  • verification – AI-generated output must not be considered reliable merely because it was generated by an AI system;
  • confidentiality – confidential manuscripts and review materials must not be submitted to external generative AI systems;
  • academic integrity – AI must not be used for fabrication, falsification, plagiarism or deliberate misrepresentation.

2. AI Cannot Be an Author

Artificial intelligence tools, chatbots, large language models and other automated systems cannot be listed as authors or co-authors of an article.

Authorship entails responsibility for the content of the work, participation in the scholarly process and the ability to accept responsibility for published statements. An AI system cannot fulfil these functions.

The use of AI does not relieve any author of responsibility for the content of the submitted or published work.

3. Use of AI by Authors

Authors may use AI tools as supporting instruments during research and manuscript preparation, provided that such use does not violate academic integrity, confidentiality, copyright or other applicable requirements.

Permitted uses may include, among others:

  • checking spelling, grammar, punctuation and style;
  • language editing and improving readability;
  • assistance with translation, provided that the result is checked by the author;
  • structuring or preliminary organisation of material;
  • assistance with programming code or scripts;
  • supporting classification, grouping or preliminary processing of large datasets;
  • generating ideas or questions for further independent scholarly work by the author.

Such use does not change the author’s responsibility for the scholarly content of the work.

4. Uses of AI Requiring Disclosure

Authors must disclose the use of generative AI where such use goes beyond ordinary language correction, editing or formatting and has materially contributed to the preparation of the manuscript.

Disclosure is required, in particular, where AI is used for:

  • generating or substantially rewriting text;
  • developing a structure or argument that is subsequently used in the manuscript;
  • analysing, classifying or interpreting datasets;
  • writing or substantially generating code used as part of the research methodology;
  • translating substantial parts of the manuscript where the translation materially shapes the final text;
  • creating tables, figures, images or other content;
  • other uses that may have materially influenced the preparation or content of the publication.

The mere use of AI is not a ground for rejection where the use is appropriate, transparently disclosed and otherwise complies with the Journal’s requirements.

5. Uses of AI Not Normally Requiring Separate Disclosure

A separate AI disclosure statement is normally not required for ordinary tools used for:

  • spelling and grammar correction;
  • basic stylistic editing;
  • routine text formatting;
  • technical formatting checks.

Where an author uses generative AI for more extensive language editing or another form of assistance that goes beyond ordinary proofreading, disclosure is recommended.

6. Prohibited or Unacceptable Uses of AI

AI must not be used for:

  • fabricating or falsifying data, documents, sources, court decisions, legal acts, statistics or research findings;
  • creating nonexistent bibliographic references or quotations;
  • presenting generated material as the author’s original research without required disclosure;
  • plagiarism or concealed appropriation of another person’s material;
  • deliberately misleading readers about sources, methodology or findings;
  • violating confidentiality or third-party rights.

Authors must independently verify all factual claims, references, quotations, legal provisions, case law and other results obtained with the assistance of AI.

Generative AI cannot replace verification of primary sources.

Generative AI must not be treated as an authoritative scholarly source on which a substantive conclusion is based. Where a claim has a primary source, the author should cite that primary source.

7. AI Use Disclosure Statement

Where disclosure is required, the statement should be included in the manuscript in the location specified by the Journal, preferably before the reference list or in another location appropriate to the structure of the article.

Recommended format:

“During the preparation of this article, the author(s) used [name of tool / model] for [description of the purpose and nature of use]. The author(s) independently verified the resulting output, factual information, references and quotations and remain fully responsible for the content of the publication.”

Where relevant, authors are also encouraged to provide the tool version or date of use, if such information is available.

8. Reviewers and Confidentiality

Reviewers must not upload a submitted manuscript or parts of it to generative AI systems where doing so could disclose confidential manuscript content or personal information about participants in the editorial process.

Reviewers must not use generative AI to generate a peer review, assess a manuscript or formulate a recommendation concerning acceptance or rejection.

Reviewers must not use unpublished ideas, data, arguments or other information from a manuscript through AI tools or in any other manner for their own purposes.

Basic language-checking tools may be used on a reviewer’s own review text only where such use does not reveal confidential manuscript content.

9. Editors and the Editorial Board

Editors and members of the editorial board must not use generative AI to automatically generate manuscript assessments, editorial recommendations or final decisions concerning acceptance or rejection.

Publication decisions are made by humans. Generative AI cannot make or determine an editorial decision.

The Journal may use specialised automated tools for technical and integrity-related checks, including screening for text similarity, duplication, metadata issues or other potential integrity concerns.

The results of such tools are not by themselves sufficient grounds for rejecting a manuscript. Automated results must be reviewed by a human and assessed in context.

10. Automated Integrity and Technical Checks

The Journal may use automated tools to support technical checks of manuscripts, metadata, possible text overlap, duplication and other relevant characteristics of submitted materials.

Such tools may include algorithmic or AI-based components.

The Journal does not rely exclusively on automated conclusions. Any potentially problematic result must be reviewed by an editor or other authorised person.

In particular, suspected AI use by an author must not be established solely on the basis of an automated AI-detection score. Decisions concerning academic integrity must be based on a combination of verifiable facts and evidence.

11. Authors’ Responsibility

Authors remain fully responsible for the accuracy, originality, legality and scholarly soundness of their publication regardless of whether AI tools were used during its preparation.

Authors must:

  • independently verify AI-generated output;
  • verify all sources, quotations and references;
  • not present fabricated or unverified information as fact;
  • comply with copyright and confidentiality requirements;
  • disclose substantive use of generative AI as required by this Policy.

Attributing an incorrect statement, fabricated reference or other inaccuracy to an AI tool does not relieve the author of responsibility.

12. Breaches of this Policy

Where a material breach of this Policy is established, the Journal may take action proportionate to the nature and consequences of the breach.

Possible actions include:

  • requesting additional information or clarification from the author;
  • requiring correction or completion of the AI disclosure;
  • rejecting the manuscript;
  • subjecting the manuscript to additional editorial or peer review;
  • for a published article, initiating the applicable correction, Expression of Concern or retraction procedure.

The appropriate response depends on the nature and seriousness of the breach and its effect on the reliability of the publication.

13. Related Policies

This Policy operates together with:

The Journal periodically reviews this Policy in light of technological developments and changes in responsible scholarly publishing practice.