AI Disclosure Requirements have become a practical submission issue for collaborative publishing projects, especially where several disciplines, methods, and writing habits meet in one manuscript. For authors, editors, and project leads, the question is no longer only whether an AI tool was used. The working question is who used it, for what purpose, how the output was checked, and where that information belongs in the submission package.
This matters because collaborative work can spread responsibility across many hands. A statistician may use software support during analysis notes. A humanities scholar may test phrasing during revision. A biomedical coauthor may use a tool to organize a figure caption. A project coordinator may not see each step unless the team asks early and records answers in a shared place.
As Leela Sharma, I view this as a publishing process issue rather than a technology trend. Disclosure is now part of manuscript management, rights awareness, authorship accountability, and editorial trust. It should be treated with the same care as contributor roles, funding statements, conflict-of-interest forms, data availability notes, and permissions records.
Why AI Disclosure Requirements Matter In Collaboration
AI Disclosure Requirements At Project Start
For a cross-disciplinary group, AI Disclosure Requirements should be discussed before drafting begins, not after the manuscript is ready to upload. The reason is practical: each field may bring different assumptions about acceptable writing help, statistical support, image handling, translation, or formatting. If the team waits until submission, it may be difficult to reconstruct who used which tool and what human checks were applied.
Two recent policy examples show why early planning is sensible. In August 2026, Elsevier released updated generative AI policies for its journals. Those policies require authors to submit a separate AI declaration statement at manuscript submission, identifying the AI tools used, the purpose of use, and the extent of human oversight. Elsevier also stated that image generation or editing must be disclosed in captions and in the general AI statement where relevant Elsevier AI policies.
The CDC published its scientific-work disclosure considerations on May 28, 2026. It stated that substantive use of generative AI in text, image, or data-analysis work should be clearly disclosed, including the tool name, version, purpose, and extent of human oversight. The CDC also stated that AI cannot be listed as an author CDC disclosure considerations.
Why Interdisciplinary Work Raises The Stakes
Interdisciplinary projects often combine different evidence standards. A public health article may need clear methods reporting. A digital humanities project may include text processing and interpretive commentary. An engineering collaboration may include data modeling, visual materials, and technical writing. If one contributor uses an AI system to summarize background literature, another uses it to revise plain-language sections, and a third uses it to adjust an image, one blanket sentence may not be enough for the target journal.
The better approach is to treat disclosure as a shared record. That record does not need to be dramatic or promotional. It should be plain, specific, and checkable. It should also separate minor tools from substantive use, because some policies distinguish between basic grammar or spelling assistance and generative work that affects content, structure, analysis, or visual materials.
What Authors Need To Record Before Submission
Build A Project-Level AI Use Log
A shared AI use log can reduce confusion before journal submission. It should be simple enough that coauthors will use it, but specific enough to support an accurate declaration. The log can sit with other project documents, such as author contribution notes, permissions files, dataset records, ethics approvals, and correspondence about figures or supplementary material.
At minimum, the log should help the corresponding author answer the questions now appearing in publisher and institutional guidance: which tool was used, which version if known, who used it, when it was used, for what task, what material went into the tool, what output was retained, and how human authors checked the result. For images or captions, the record should identify whether the tool generated, edited, enhanced, or reorganized visual content.
- Tool and version: Record the name and version where available, rather than using a vague phrase such as “AI was used.”
- Purpose: State whether the tool supported drafting, language editing, coding support, data analysis, image work, formatting, or another task.
- Human oversight: Identify who reviewed, corrected, verified, or rejected the output.
- Manuscript location: Note whether the use affects methods, acknowledgments, figure captions, supplementary material, or a separate declaration form.
- Submission owner: Assign one person to compare the log with the target journal’s instructions before upload.
Keep Authorship Separate From Tool Use
Journal and institutional policies cited in the research notes share a central principle: human authors remain responsible for the manuscript. A tool cannot approve the final article, take responsibility for accuracy, answer editorial questions, handle corrections, or accept accountability for citations and claims. That is why AI use records should support human responsibility rather than replace it.
This distinction is especially important in collaborative works where senior, junior, and external contributors may have different levels of control over the final file. If a graduate researcher, contractor, editor, translator, or technical contributor uses a tool, the named authors still need a reliable way to verify the material before submission. That does not mean every project needs a long statement. It means the team should avoid guessing.
Authors who are already tracking editorial use of AI may also want to compare disclosure notes with submission-stage policy checks. A related discussion of AI peer review policies explains why manuscript handling rules can affect authors before and after acceptance.
How Interdisciplinary Teams Can Set A Shared Process
Assign Responsibility Without Creating Bottlenecks
AI Disclosure Requirements should be assigned to a person, but the work should not depend on that person discovering everything alone. A corresponding author, project manager, or lead editor can own the final declaration, while each contributor confirms their own tool use. This keeps the process accountable without asking one person to audit private drafting habits after the fact.
A useful workflow is to add three checks to the manuscript calendar. First, ask about AI use when contributors accept their roles. Second, update the log after major drafting, analysis, figure, or revision milestones. Third, review the target journal’s policy before submission, because publisher requirements can differ in wording and placement.
For edited collections, conference proceedings, and multi-author special issues, the volume editor may need a separate process. Chapter authors may submit work at different times and under different disciplinary norms. A short contributor form can help collect consistent information, especially when the final publisher asks for one declaration across the project.
Connect Disclosure With Contracts And Permissions
AI disclosure is not a substitute for contract review, copyright clearance, data protection, or ethics approval. It does, however, sit near those records. For those interested in comparing how credibility, documentation, and service descriptions vary across professional fields, the contrasting approaches offered by the Polygraphianz site can offer valuable insights next to academic publishing, where journal instructions and author accountability set the main terms.
Marketing and public communication should stay aligned with the disclosure record. If a project uses AI for a plain-language summary, promotional copy, or a visual abstract after acceptance, that may not be part of the original manuscript declaration, but it can still raise questions about accuracy, attribution, and rights. Teams should decide who reviews public-facing material before it goes live.
AI Disclosure Requirements For Interdisciplinary Works
Use Plain Language And Avoid Overclaiming
AI Disclosure Requirements will not make every collaborative project simple. They can, however, make responsibility easier to see. The strongest statements are usually direct: name the tool, describe the task, state who checked the output, and place the disclosure where the journal asks for it. Avoid broad claims that a tool “improved quality” unless the journal specifically asks for evaluative wording and the team can support the claim.
For interdisciplinary authors, the best preparation is ordinary publishing discipline. Keep records while the work is happening. Match the declaration to the target journal’s instructions. Treat image use separately when policy requires it. Do not list a tool as an author. Do not rely on AI-generated references without human verification. Make sure every named author has a chance to review the final statement.
The practical benefit is not only compliance. Clear disclosure helps editors understand the manuscript file they are receiving, helps coauthors agree on accountability, and helps readers trust that human experts stand behind the work. In collaborative publishing, that trust is built before submission, one documented decision at a time.