Artificial intelligence (AI) is becoming an increasingly common part of academic and scientific writing. Researchers may use AI tools to brainstorm ideas, improve language, summarize literature, organize content, analyse information, or refine a manuscript.
However, using AI in academic writing comes with an important responsibility: transparency.
Universities, journals, publishers, and research organizations are increasingly asking authors to disclose when and how AI has contributed to their work. The purpose is not to discourage responsible AI use. Rather, disclosure helps preserve research integrity, authorship accountability, and reader trust.
For researchers preparing manuscripts, knowing when AI disclosure is required—and how to document it correctly—is therefore essential.
Why Does AI Disclosure Matter in Academic Writing?
Academic publishing depends on transparency. Readers, reviewers, and editors need to understand how research and manuscripts were developed and who is ultimately responsible for the content.
When AI has made a meaningful contribution to a manuscript, openly acknowledging its use demonstrates that the authors have reviewed and taken responsibility for the final work.
AI tools can assist with writing and research-related tasks, but they cannot take responsibility for scientific claims, interpretations, references, data, or conclusions. Human authors remain accountable for the accuracy and integrity of the submitted work.
Undisclosed or inappropriate AI use can also create concerns about:
- Research integrity and academic misconduct
- Accuracy of AI-generated information
- Fabricated or incorrect references
- Loss of the author's original voice
- Questions about authorship and intellectual contribution
- Compliance with journal or institutional policies
The key principle is simple: AI can support the research and writing process, but responsibility remains with the human author.
When Should You Disclose AI Use?
AI use should generally be disclosed when the tool has made a meaningful contribution to the development of your manuscript.
Examples include:
- Generating an initial outline or structure
- Drafting or substantially rewriting sections of text
- Summarizing research literature
- Assisting with data analysis or coding
- Creating or modifying figures, tables, or other visuals
- Translating or substantially editing manuscript content
- Helping interpret or organize information used in the manuscript
For example, if an AI tool is used to summarize several published studies and those summaries influence the development of your literature review, that contribution should be disclosed.
What About Minor AI Assistance?
Not every interaction with an AI-powered tool necessarily requires the same level of disclosure.
Basic spelling corrections, punctuation checks, or routine grammar assistance may be treated differently from generating paragraphs, restructuring arguments, or analysing information. However, requirements vary between journals and institutions.
When in doubt, transparency is the safer approach.
Always review the specific AI and author guidelines of your target journal before submission.
Where Should AI Use Be Disclosed?
The appropriate location for an AI disclosure depends on the journal, publisher, and nature of the AI assistance.
Common locations include:
Methods or Methodology
If AI was used as part of the research process—for example, data analysis, coding, modelling, or another technical procedure—the use may need to be described in the Methods section.
Acknowledgements
AI assistance involving language editing, writing support, or other non-research contributions may be acknowledged here when permitted by the journal.
Dedicated AI Disclosure Section
Some publishers request a specific statement such as “Declaration of Generative AI and AI-Assisted Technologies.”
Cover Letter
A journal may also ask authors to disclose AI use in the cover letter accompanying the manuscript.
Because requirements differ, authors should check the latest instructions for authors before submission.
What Should an AI Disclosure Statement Include?
A strong disclosure should be specific rather than vague.
Where applicable, include:
- Name of the AI tool – such as ChatGPT or another AI-assisted platform.
- Purpose of use – explain what the tool was used for.
- Scope of involvement – identify the manuscript section or research activity affected.
- Timing – include the relevant date or period when appropriate.
- Human review – confirm that the authors reviewed, verified, and edited the AI-assisted output.
- Author responsibility – clearly state that the authors remain responsible for the final manuscript.
Avoid statements such as:
“AI tools were used in preparing this manuscript.”
This is too broad to explain what the AI actually did.
A more useful statement would be:
“The authors used [AI Tool] to assist with language editing and the initial organization of background literature. All AI-assisted content was reviewed, fact-checked, and revised by the authors, who retain full responsibility for the accuracy and integrity of the manuscript.”
The exact wording should always be adapted to the journal's requirements and the actual use of AI.
Can AI Be Listed as an Author?
No.
AI systems cannot assume responsibility for a manuscript and therefore should not be listed as authors.
Authorship carries responsibilities relating to the accuracy, integrity, interpretation, and accountability of published research. These responsibilities can only be assigned to human contributors who meet the relevant authorship criteria.
AI should therefore be treated as a tool or technology used during the research and writing process—not as a co-author.
Common Mistakes to Avoid
Researchers should be particularly careful about the following:
1. Hiding meaningful AI use
If AI substantially contributed to the manuscript, failing to disclose it can create questions about research integrity.
2. Assuming AI-generated information is accurate
AI systems can produce incorrect facts, fabricated references, misleading interpretations, or outdated information. Every AI-assisted output must be independently verified.
3. Allowing AI to replace scholarly judgment
AI can help organize and develop ideas, but researchers should remain responsible for evaluating evidence, interpreting findings, and drawing conclusions.
4. Using vague disclosure language
Simply saying “AI was used” does not tell readers how the technology contributed to the work.
5. Ignoring journal-specific requirements
AI policies are evolving rapidly. Requirements can differ significantly between publishers and individual journals.
A Practical Step-by-Step Approach
A simple workflow can help researchers manage AI use responsibly:
Step 1: Record how AI was used.
Keep track of whether AI assisted with drafting, editing, summarization, analysis, coding, translation, or other activities.
Step 2: Assess the level of contribution.
Determine whether the AI use was routine and minor or meaningfully influenced the manuscript.
Step 3: Verify all AI-assisted content.
Check facts, references, statistics, interpretations, and scientific claims against reliable primary or authoritative sources.
Step 4: Prepare an appropriate disclosure.
Clearly state what tool was used, why it was used, and how the authors reviewed its output.
Step 5: Check the target journal's guidelines.
Review the journal's current instructions on generative AI, authorship, disclosure, and research integrity before submission.
Step 6: Keep appropriate records.
Where practical, retain prompts, drafts, revisions, and relevant documentation of AI involvement. This can help demonstrate responsible authorship if questions arise.
Transparency Is the Best Practice
The debate around AI in academic publishing is likely to continue as technologies and policies evolve. What remains consistent is the importance of transparency and human accountability.
Responsible AI use does not mean allowing technology to take over the intellectual work of the researcher. Instead, AI can be used as a supporting tool while researchers retain control over the scientific reasoning, evidence evaluation, interpretation, and final manuscript.
For authors preparing research for publication, the safest approach is straightforward:
Use AI responsibly. Verify everything. Disclose meaningful use. Follow the target journal's guidelines. Take full responsibility for the final work.
At Corriger Global, we believe that effective scientific and academic communication must combine technological innovation with research integrity. As AI becomes more integrated into scholarly workflows, clear disclosure and responsible human oversight will remain fundamental to credible scientific publishing.