03 · What You Need to Know
How to Choose the Appropriate Location for AI Disclosure
Why Does the Location of an AI Disclosure Matter?
Different manuscript sections serve different scholarly purposes.
The methods section explains how research was conducted. It provides information needed to evaluate the design, procedures, analysis, and limitations of a study.
The acknowledgments section ordinarily recognizes assistance that supported the work without necessarily qualifying a contributor for authorship.
A dedicated AI-use declaration may serve a publisher's administrative and transparency requirements without replacing methodological reporting.
These functions can overlap, but they are not identical.
Methods section
Explains AI-assisted procedures that materially affected how research data were collected, processed, analyzed, or interpreted.
Acknowledgments section
Recognizes AI assistance with manuscript preparation where this is an appropriate reporting location under the journal's policy.
Choosing the correct location helps readers distinguish between technology used to conduct the research and technology used to communicate it.
What Does the ICMJE Recommend?
The ICMJE recommends that authors disclose whether AI-assisted technologies were used in producing submitted work and explain how they were used.
Its guidance distinguishes between AI assistance with writing and AI involvement in research activities.
Writing assistance is to be reported in the acknowledgment section, while AI use for data collection, analysis, or figure generation should be described in the methods section as applicable.
The ICMJE also addresses disclosure in submission correspondence, so reporting AI use within the manuscript may not satisfy every submission requirement.
These recommendations are particularly influential in medical publishing, although journals in other fields may adopt them.
Researchers should consult the journal's current instructions rather than assuming that every publisher follows the ICMJE reporting structure.
What Does COPE Recommend?
The Committee on Publication Ethics (COPE) states that AI tools cannot be listed as authors and that authors must remain responsible for AI-assisted content.
Its position on authorship and AI tools also emphasizes transparency about AI use in manuscript writing, image production, and research activities.
COPE specifies disclosure of the tool and how it was used in materials and methods or an equivalent section.
This differs from the ICMJE's explicit distinction between acknowledgments for writing assistance and methods for methodological AI use.
The positions are consistent in requiring transparency and human accountability, but they do not prescribe identical manuscript locations.
A journal's own author instructions should therefore determine how these principles are implemented in a particular submission.
How Does Springer Nature Approach AI Reporting?
Springer Nature distinguishes different categories of AI assistance and provides reporting instructions according to their role.
Its guidance permits certain declarations in the introduction or acknowledgments while addressing methodological AI use through more specific reporting requirements.
It also distinguishes routine AI-assisted copyediting from substantive generative AI assistance. Under its defined conditions, some copyediting activities do not require declaration.
These distinctions demonstrate why the reporting location cannot be determined solely from the fact that an AI tool was used.
The researcher must identify the actual activity and consult the applicable publication policy.
How Do the Main Reporting Approaches Compare?
| Publishing Authority |
Writing Assistance |
Methodological AI Use |
| ICMJE |
Acknowledgments, with applicable submission disclosure. |
Methods section for data collection, analysis, or figure generation. |
| COPE |
Materials and methods or equivalent section, according to its position statement. |
Materials and methods or equivalent section. |
| Springer Nature |
Introduction or acknowledgments may be appropriate; defined copyediting exceptions apply. |
Follow methodological reporting and journal-specific requirements. |
This comparison summarizes general positions rather than every policy applicable to journals associated with these organizations.
Researchers should verify the current instructions for the publication receiving their manuscript.
When Does AI Use Become a Methodological Issue?
AI use becomes methodologically relevant when it contributes to procedures through which research evidence is generated, processed, selected, analyzed, or interpreted.
For example, an AI system might assist with classifying survey responses, extracting information from documents, generating analytical code, or identifying themes in interview transcripts.
These activities may affect the findings and therefore require sufficient information for readers to understand how the research was conducted.
A brief acknowledgment naming the tool would not necessarily explain the procedure or allow readers to evaluate its limitations.
Methodological reporting should describe the AI contribution at a level appropriate to its significance and the journal's standards.
Should AI-Assisted Qualitative Coding Be Described in Methods?
Generally, yes, when AI materially contributes to the coding or interpretation of research data and the journal requires methodological transparency.
Qualitative coding involves analytical decisions about how data are categorized and interpreted.
If AI suggests codes or themes, readers may need to know how those suggestions were generated, reviewed, modified, and incorporated.
For example, a researcher might use an AI model to propose preliminary codes and then conduct a human review of every coded excerpt.
The methods section should accurately describe that process rather than merely stating that AI assistance was acknowledged.
Questions about disclosing AI-assisted qualitative analysis are particularly important because the tool may influence interpretive judgments rather than only manuscript wording.
Should AI-Generated Statistical Code Be Reported in Methods?
When AI-generated code materially contributes to the analysis, methodological reporting may be appropriate or required.
The relevant information may include the analytical procedure, software environment, code verification, and role of the AI tool.
The extent of detail depends on the research design and publication requirements.
Researchers should distinguish between AI assistance used to write code and the statistical method actually implemented.
A manuscript must accurately report the analysis performed, regardless of how the code was generated.
The question of disclosing AI-generated research code also involves reproducibility and verification.
What If AI Was Used Only to Improve Manuscript Language?
Language assistance generally concerns manuscript preparation rather than the underlying research methodology.
Where disclosure is required, acknowledgments or a designated AI-use statement may therefore be appropriate.
However, journal policies differ in how they treat grammar correction, translation, paraphrasing, and substantive rewriting.
A tool used only to correct punctuation may fall under an exemption, while a tool used to generate or reconstruct arguments may require more detailed disclosure.
The researcher should accurately describe the assistance and avoid presenting substantive drafting as routine copyediting.
What If the Same AI Tool Was Used for Both Writing and Analysis?
A single AI tool can contribute to several distinct research activities.
For example, ChatGPT might be used to suggest analytical code and later to improve the wording of the discussion.
The tool's identity does not determine the disclosure location. Its role does.
Under a policy distinguishing writing and methodological assistance, the researcher may need to describe the analytical use in methods and report writing assistance in acknowledgments or another designated section.
These statements should be consistent without unnecessarily repeating the same information.
Where the journal requires one consolidated AI-use declaration, researchers should still ensure that any methodological information needed for reproducibility appears in the appropriate research section.
What Details Belong in a Methodological AI Disclosure?
Methodological reporting should provide enough information to understand the AI-assisted procedure and evaluate its implications for the findings.
Depending on the activity, relevant details may include:
- The AI tool or model used and its role in the procedure.
- The types of inputs or research materials processed.
- The tasks assigned to the system and how outputs were incorporated.
- The human verification, correction, or validation performed.
- Relevant limitations and reproducibility information.
Some research contexts may also require model versions, dates, settings, or prompt documentation.
These details should be proportionate to the methodological significance of the AI assistance.
The question of what an AI disclosure statement should contain should be distinguished from the fuller reporting required for reproducible research methods.
Can an Acknowledgment Replace a Methods Description?
Not when the methods description is necessary to understand or evaluate the research.
An acknowledgment can identify that assistance occurred, but it may not explain how AI affected sampling, coding, analysis, or interpretation.
For example, a statement that an AI model assisted with thematic analysis would not necessarily reveal whether it independently generated codes, classified all transcripts, or merely suggested possible categories for human consideration.
Those differences may materially affect how readers interpret the findings.
Methodological transparency therefore requires reporting the actual procedure, not merely naming the tool.
Does Reporting AI in Methods Make Its Use Automatically Acceptable?
No. Disclosure and permission are separate questions.
A journal may restrict particular AI activities even when they are fully documented.
Researchers must therefore establish whether the assistance is permitted and whether it has been reported appropriately.
They also remain responsible for the validity of the resulting research.
A detailed methods disclosure does not excuse unverified analysis, inappropriate processing of confidential data, or unsupported interpretations.
What If the Journal's Instructions Are Ambiguous?
Researchers should review the journal's author guidelines, publisher-level AI policy, and submission declarations.
If the appropriate reporting location remains unclear, they may request clarification from the editorial office.
Any clarification should be retained with the manuscript preparation records where appropriate.
Researchers should not assume that a disclosure placed somewhere in the manuscript automatically satisfies an unspecified requirement.
Clear reporting is preferable to a technically present but insufficiently informative statement.
