AI Use Policy
This policy covers AI tools, including large language models and generative assistants, used in preparing or reviewing a manuscript. It follows COPE guidance. Because many JHAI papers develop or evaluate clinical AI, the policy separates two things: AI used as a writing or preparation aid, which the rules below govern, and AI as the subject of the research, which is described in the Methods like any other technique.
1. AI cannot be an author
AI tools do not meet the authorship criteria. They cannot approve a manuscript or be accountable for it, so they must not be listed as authors. The human authors are responsible for everything in the manuscript, including any part produced with AI help.
2. Permitted uses
- language editing and grammar checking;
- code completion during software development, with the authors reviewing all generated code;
- help with literature searches, provided every citation is checked by the authors;
- summarising or rewording the authors' own earlier text, which the authors then review and edit;
- AI as a method or model under study, fully described in the manuscript.
3. Prohibited uses
- submitting AI-generated text as the authors' own without review and full responsibility;
- references invented by AI (authors must check that every reference exists);
- AI-generated data, results or images presented as real;
- altering figures or images with AI in a way that misleads;
- paraphrasing copied text with AI to avoid plagiarism detection.
4. Disclosure
Any use beyond basic language editing must be disclosed in a "Use of AI Tools" statement or a subsection of the Methods, giving:
- the tool and version (for example ChatGPT, Gemini, Claude or Copilot);
- what it was used for;
- which parts of the manuscript were affected;
- a statement that the authors take full responsibility for the content.
"During the preparation of this manuscript the authors used ChatGPT (OpenAI) to improve the language of the Introduction and Discussion. The authors reviewed and edited the output and take full responsibility for the content of the publication."
This disclosure is not needed for AI that is itself the object of study. That is reported in the Methods.
5. Figures and images
AI may not be used to create or alter figures, except where the AI-generated content is the subject of the research and is labelled as such. AI-generated schematic illustrations must be declared in the caption.
6. Reviewers and editors
Reviewers and editors must not upload any part of a manuscript under review to an AI service. They may use AI to polish the language of their own report, provided no manuscript content is shared.
7. Checks
The journal may ask for source code, raw outputs or AI transcripts if the authenticity of any part of a manuscript is in doubt. Undisclosed substantive AI use may be treated as misconduct under the Publication Ethics policy.
