Interactive AI Disclosure Generator

Manuscript and Author Details

Manuscript Title:

Corresponding Author Name:

Corresponding Author Email:

AI Use Date Range:

Start Date

End Date

List of AI Tools Used (with version numbers):

Target Journal Name:

Conceptual Contributions

Did you use AI for this domain?

Yes No

Check one of the boxes in the two columns - generally acceptable or "May not be acceptable". Then add any comments if necessary to explain your particular use of AI tools:

Idea Generation and Refinement

Generally Acceptable: Author/s generated the initial idea/s or hypothesis and used AI to brainstorm refinements.

May not be acceptable: Asked AI to originate the initial idea or hypothesis without significant prior human thought process, research or discussion.

Comments:

Structure/flow

Generally Acceptable: Author/s came up with the initial outline, flow and structure (including section headers and paragraphs) of the manuscript before asking AI for suggestions. Author/s reviewed the AI input and edited it appropriately before applying it.

May not be acceptable: Asked AI to create the initial outline, flow and structure of a manuscript. Using the AI output without human review or edits.

Comments:

Argument Testing and Critical Analysis

Generally Acceptable: Asking AI to identify weaknesses, counter arguments or missing perspectives on a near final draft of the manuscript.

May not be acceptable: Asking AI to write the limitations, discussions or conclusion sections of a manuscript without an initial human draft.

Comments:

Summarization/ Synthesis

Generally Acceptable: Using AI to condense complex information from multiple sources, identify key themes, or synthesize findings with human verification of accuracy and review of selection of content being summarized.

May not be acceptable: Accepting AI selection of content, summarization and synthesis without human review.

Comments:

Conceptual Analogies or Frameworks

Generally Acceptable: Asking LLMs to propose metaphors or conceptual models to explain complex ideas, followed by human validation and adaptation.

May not be acceptable: Using AI generated models or metaphors without ensuring that they fit the intended message of the manuscript.

Comments:

Visual Abstracts and Infographics

Generally Acceptable: Utilizing AI to draft visual abstracts or infographics based on manuscript content, with subsequent human review, editing, and factual verification and attribution to AI.

May not be acceptable: Accepting AI generated visual content without human review and verification of accuracy or copyrighted content or style.

Comments:

Linguistic Assistance

Did you use AI for this domain?

Yes No

Check one of the boxes in the two columns - generally acceptable or "May not be acceptable". Then add any comments if necessary to explain your particular use of AI tools:

Language Editing

Generally Acceptable: Using AI for grammar, spelling, punctuation.

May not be acceptable: Accepting edits without human review, potentially altering scientific meaning.

Comments:

Clarity, Conciseness and Readability

Generally Acceptable: Using AI for sentence-level editing to refine phrasing and syntax, streamline wording, and organize sentence/paragraph structure so the text is clear, concise, and easy to read, without changing meaning with subsequent human review and editing.

May not be acceptable: Using AI edits without human verification for accuracy, shifts in scientific meaning or terminology, or additions/removals of essential content.

Comments:

Language Translation for non-native speakers

Generally Acceptable: Using AI to generate a translation draft with subsequent review by bilingual co-author, subject-matter expert or professional translator.

May not be acceptable: Using the AI generated translation without a human review.

Comments:

Stylistic Consistency

Generally Acceptable: AI may be used to standardize to a journal’s style and formatting per a specified guide, but authors must review changes, ensure no scientific meaning or data are altered.

May not be acceptable: Using the AI generated manuscript changes without human review and editing.

Comments:

Adjusting for required Word Count

Generally Acceptable: AI may be used to modify the abstract or article to meet a journal’s word count, but authors must review changes, ensure no scientific meaning or data are altered.

May not be acceptable: Using the AI generated manuscript changes without human review and editing.

Comments:

Suggesting Title Variations or Subheadings

Generally Acceptable: AI may be used to suggest variations of titles or sub-headings that humans choose from.

May not be acceptable: Accepting the AI suggestion without human review to ensure alignment with the manuscript.

Comments:

Research Assistance

Did you use AI for this domain?

Yes No

Check one of the boxes in the two columns - generally acceptable or "May not be acceptable". Then add any comments if necessary to explain your particular use of AI tools:

Research Question Refinement

Generally Acceptable: Clarify constructs e.g. help draft the specific PICO question but author/s select and justify the final specifications.

May not be acceptable: Delegating question selection or outcome specification to AI without human review.

Comments:

Search Strategy and Strings

Generally Acceptable: Draft Boolean strings, MeSH terms and filters for PubMed search validated by author/s or information specialist.

May not be acceptable: Using AI generated search strategies without human or expert review.

Comments:

AI-Assisted Literature Retrieval and Aggregation

Generally Acceptable: Use tools like Elicit and Undermind to propose candidate papers and clusters for human screening or supplementary inputs. Author/s information specialists use documented databases for searches for missed articles.

May not be acceptable: Accepting AI generated literature search results and summaries without searching documented databases.

Comments: