AI-Use Disclosure Statement: How to Write One (Templates for Students and Researchers)

Published August 5, 2026 · FiftyGPT Editorial Team · 11 min read · Updated October 11, 2026

To write an AI-use disclosure statement, name each AI tool that materially shaped the work, state what it did and where, identify retained output, describe only the checks you actually ran, and accept responsibility. Then match that account against your course or publisher policy. Below: a six-field record, a decision tree, three templates, and a checklist.

What is an AI-use disclosure statement?

An AI-use disclosure statement is a short factual record explaining how AI tools contributed to a piece of submitted work — written for an instructor, supervisor, editor, or reviewer. It reports what happened. It does not make prohibited assistance permissible, and it is not an apology.

Detection cannot replace this documentation. The ethics of AI detectors explains why detector scores are probability signals rather than proof of process, and the guide on talking to students suspected of AI use shows why a written record resolves disputes faster than any score.

This matters because AI use in research is now the norm, not the exception. Nature's 2023 survey of more than 1,600 researchers found AI tools becoming central to scientific practice (Nature, 2023), and a study of US academic scientists found 65% had personally used generative AI (PMC, 2025). Even peer review isn't exempt: a 2025 survey of 1,600 academics found more than half had used AI tools while peer reviewing manuscripts (Nature, 2025). When everyone uses these tools, the differentiator is who documents it honestly.

Note the boundary: the method below is editorial drafting guidance, not universal academic law. Confirm the actual requirement in your assignment brief, academic integrity policy, thesis regulations, or journal author guidelines before relying on any wording here.

How do you start? Write a six-field factual record first

Write a private record before a polished sentence. Polished wording can read as complete while hiding a missing fact. Create one entry for every tool-task pair that materially influenced the submitted work:

1. Tool: the product used. Record the version or model only if known and asked for — never guess.
2. Task: a precise verb — brainstormed, outlined, drafted, rewrote, corrected, translated, summarized, analyzed, generated code.
3. Scope: the affected chapter, paragraph, figure, dataset, calculation, or stage.
4. Retained output: whether generated ideas, wording, translations, summaries, analysis, or code remain in the submission.
5. Verification: a check you genuinely completed — not the one that sounds most reassuring.
6. Responsibility: confirmation that you reviewed and accept responsibility for the work.

Worked example. A student asked a tool for four possible essay headings, kept two after revising them, wrote the paragraphs independently, and checked the structure against the assignment question. The record converts to: "I used [tool] to suggest four essay headings. Two revised headings remain; I wrote the paragraphs independently. I checked the structure against the assignment question and take responsibility for the submission."

Notice what the example avoids: "the tool helped with planning" hides the retained headings; "the tool wrote the essay" overstates its role. Precision is the entire point.

For the editing side of this workflow, the free grammar checker handles mechanical correction — and knowing exactly what it did makes the record easier to write. The academic tools directory collects the planning and checking tools researchers use alongside disclosure.

Which AI uses require disclosure? A decision tree

Classify the involvement into five categories. The tree names the contribution precisely; it cannot decide whether the contribution was permitted — only your policy does that.

• Ideation (brainstorming): did the tool shape the research question, angle, examples, or structure? A retained outline can be material even when every final sentence is your own. Rule: disclose when a generated idea or structure survives into the submission.
• Drafting: did the tool produce submitted sentences, paragraphs, captions, or an abstract? Rule: always disclose retained generated prose and name where it sits. Editing it later doesn't turn drafting into ideation.
• Editing or grammar: separate mechanical correction from substantive rewriting. Fixing isolated spelling, punctuation, or agreement is mechanical; changing tone, argument order, or substantial wording is rewriting. Rule: mechanical correction is often exempt — confirm against the policy — but rewriting normally requires disclosure.
• Translation: did the tool create submitted text in another language? Record source and target languages, affected sections, and the meaning check you ran. Rule: disclose whenever translated text is submitted, because a fluency read is not a comparison with the source.
• Data analysis or code: did the tool propose a method, calculation, query, or script? Name the component and the inspection, tests, or comparisons you completed. Rule: disclose any retained method or code, and never claim comprehensive testing after running it once.

When one tool did materially different jobs — suggesting an outline and later rewriting a conclusion — give each its own clause. Combining them as "writing support" hides the drafting contribution.

If you're unsure where your use falls, the AI detector FAQ explains what detection can and can't establish, and how AI detectors work clarifies why a detector score is not a substitute for your own record.

How much detail should you include? The material-detail test

Include every tool-task pair that materially influenced the work — its scope, retained output, and completed checks. Leave out experiments that changed nothing. Don't pad with prompt-by-prompt history unless the policy asks for a full log. Three questions decide:

1. Would the submission be meaningfully different without this interaction? An outline that reordered your argument is material even if no generated sentence remains. A rejected response that influenced nothing is immaterial — unless all use must be reported.
2. Does the wording conceal anything? "I used AI to improve my paper" could mean punctuation fixes, rewritten arguments, or drafted analysis. Replace vague verbs like "helped" or "improved" with the task, the location, the retained output, and the real check.
3. Is there padding? Give the shortest statement that leaves every material role clear. Marketing language, prompt logs, and model version numbers bury what matters.

Scenario. A researcher rejects three generated titles, translates two submitted paragraphs, and accepts seven sentence-level rewrites. Only the translation and the rewrites are material — each needs its own clause. But if one rejected title became the basis of the final title, that interaction turns material even though the wording changed, because materiality is about influence, not verbatim retention. The same logic applies to reworded sources — see paraphrasing to avoid plagiarism and the academic paraphrasing tool guide.

Drafting now? Run your statement through the free grammar checker for a clean mechanical pass — then record that pass in your six-field record, since precise disclosure includes your editing tools.

What should the statement actually say? Three templates

These are fill-in frameworks, not text to copy verbatim. Replace every bracket, delete any clause that isn't true, and keep the facts stable while adjusting person and placement to fit the format.

Undergraduate essay

AI-use disclosure: I used [tool] during [stage] to [task] for [scope]. [State what output remains.] I checked it by [actual method]. I reviewed and take responsibility for the final argument, evidence, citations, and wording.

Thesis or dissertation

AI-assistance declaration: I used [tool or tools] in [chapters or stages] for [specific tasks]. [Identify retained language, ideas, summaries, analysis, or code.] I checked the output by [method]. I take responsibility for the research design, source use, analysis, and conclusions.

Journal manuscript

During preparation of this manuscript, the author(s) used [tool] for [task and section]. [Describe retained output.] The output was checked using [verification method] and revised as needed. The author(s) take responsibility for the final manuscript.

Use first person only when permitted, and "the author" only where that convention fits. Never add a formal claim like "all output was independently validated" unless you can name the checks behind it. Disclosure doesn't replace source attribution either — the free citation generator helps format references as a separate step. If your thesis involved humanizing drafts, the AI humanizer for academic writing guide describes that workflow precisely, and the essay flagged as AI guide covers what to do if a draft gets questioned.

How do you disclose several tools without merging their roles?

Give every materially different tool-task pair its own row so the distinct roles stay visible. Split one product across two rows when combining its jobs would hide a difference:

Don't compress this into "I used several AI tools for research and writing" — that erases the line between translation, rewriting, summarization, and code generation. If a verification cell is blank, don't invent reassurance: run an appropriate check, or state the limitation honestly when you ask the responsible authority what's required.

Related reading: what to do if your essay is flagged as AI, fair process when AI use is suspected, and the ethics of AI detectors.

How do you match the statement to your course or publisher policy?

Before submitting, run this checklist against your specific policy language:

1. Locate the governing document — assignment brief, academic integrity policy, thesis regulations, journal author guidelines, or submission form.
2. Compare every record row with the permitted, restricted, or prohibited uses it names.
3. Use the policy's own terminology only where it accurately describes what happened.
4. Check what must be kept or submitted — prompts, outputs, drafts, tests, correspondence.
5. Place the statement where required — acknowledgment section, methods note, declaration field, title-page section, or form field.
6. Read the finished statement against your record and delete any check you didn't perform.

If assistance was prohibited, disclosure doesn't cure it. Where it was permitted, report it factually rather than apologetically. Wharton professor Ethan Mollick captures the open question well: "If you [ask AI to] write an outline for you to use, do you have to disclose that? It's hard to have a single hard-and-fast rule about when you disclose things" (Business.com interview). That uncertainty is exactly why the policy document — not your instinct — is the authority.

Also check whether universities use AI detectors and what detectors colleges actually use so your expectations about enforcement are realistic.

What if the policy is silent or ambiguous?

Keep your factual record and ask the responsible instructor, supervisor, editor, or research-office contact before submitting. Don't treat silence as automatic permission or automatic prohibition.

Send a decision-ready message: "I used [tool] to [task] in [scope]. Its output [does/does not] remain, and I checked it by [method]. The instructions don't appear to address this. Is it permitted, what must I disclose, and where should the statement go?" If one document permits general editing while another bans generated text, name both provisions — don't assume the broader phrase overrides the specific restriction. If the answer is partial, ask separately about permission, required detail, and placement; if it's verbal, keep a dated note of what was decided and when. For background on the tools themselves, see are AI humanizers safe and is the AI humanizer safe.

Disclosure is not an admission of wrongdoing

A truthful record of permitted assistance is not an admission of cheating. If the assistance was prohibited, though, disclosure doesn't make it acceptable. Disclosure records how the work was produced; AI detection does something different — it estimates patterns in finished text. Detector results are probability signals, not proof, and should be weighed with writing history, citations, drafts, and human review. If independently authored work was flagged, see how to talk to a student suspected of AI use and the fair review guide rather than treating disclosure as a defense.

The bottom line

Write the six-field record first, classify each use honestly, say only what you can verify, and match the wording to your actual policy. Disclosure takes twenty minutes; an integrity dispute takes weeks. Polish the final wording with the free grammar checker — and record that pass too. Start with the academic tools directory for the checking tools mentioned in this guide.

FAQs