Student Guide to AI Disclosure: When, How, and Templates

Search student guide to AI disclosure and the SERP is mostly institutional pages: Princeton Scholarly Integrity (disclose, don’t cite the model as a source), Oxford and Newman LibGuides, SLU/CTL disclosure PDFs, and APA/MLA citation notes. They explain that you should be transparent. Few give ready wording for the five uses students actually report — brainstorm, outline, grammar, section scaffold, paste cleanup.

This guide fills that gap: when disclosure is required, how to structure a statement, and templates you can adapt. Pair with AI writing policy for students and responsible AI writing for students. Your syllabus template wins if one exists.

When to disclose

Work top to bottom. Stop at the first match.

1. Assignment bans generative AI → Do not use it. Disclosure does not cure a ban.
2. AI produced ideas, analysis, or prose that appear in the submission → Disclose (and cite if your style guide requires it).
3. AI meaningfully supported thinking or writing — outline options, counterargument brainstorm, substantial clarity rewrites → Disclose.
4. AI was only mechanical — spell check, basic punctuation — and the syllabus exempts that → Often no disclosure; confirm Grammarly/similar carve-outs.
5. Unsure → Disclose briefly and accurately, or ask the instructor before submit.

Princeton-style guidance often separates disclosure (AI as tool/algorithm) from citation (human-authored source). MLA/APA AI citation rules are not the same as a short methods-style disclosure line. Follow your department.

How to disclose (five-part pattern)

Adapt this skeleton unless the syllabus gives a fixed form:

1. Tool and version — “ChatGPT (GPT-4o, accessed August 2026)” · “Grammarly” · “PassMyEssay”
2. Tasks — granular: brainstorm counterarguments for section 2; outline options from my notes; grammar on intro; remove invisible characters from pasted ChatGPT notes. Not “helped with essay.”
3. What you verified — citations in the library database; meaning locked after cleanup; all body paragraphs written by you.
4. Responsibility — “I am responsible for the final argument, sources, and submission.”
5. Policy anchor (optional) — “This follows the AI policy on the syllabus dated …”

Where to put it

Follow the syllabus if it specifies. Common placements:

• End of document, before references
• Title-page footnote or cover sheet
• LMS submission comment / assignment note
• Separate “AI use” appendix if required

Keep a private record: date, tool, prompt gist, affected sections. Process evidence helps if a detector later flags formal prose you wrote yourself — how to show your writing process if your essay is flagged.

Disclosure templates

Brainstorm only

"

I used ChatGPT to brainstorm counterarguments for section 2 and to clarify the assignment prompt. I wrote the full draft myself from course readings, verified all citations in the library database, and did not paste generative paragraphs into the final essay.

Outline assistance

"

I pasted my reading notes into ChatGPT and asked for two possible outline structures. I chose structure B, revised headings manually, and wrote all prose myself. Outline dated [date] available on request.

Grammar and clarity (selective)

"

I used Grammarly for punctuation suggestions on the final draft and accepted approximately ten edits to comma usage. I rejected clarity rewrites that altered claim strength. No generative drafting.

Section scaffolding (higher risk — only if policy allows)

"

I used ChatGPT to produce a rough paragraph scaffold for the methods overview from my lab notes. I rewrote every sentence in my own words, restored instrument names and sample sizes from the protocol, and verified that no citation was invented by the model.

ChatGPT paste cleanup / humanizing

"

I pasted ChatGPT-assisted notes through PassMyEssay to remove invisible Unicode watermarks and revise stock AI phrasing with meaning locked. I reviewed every change, restored any hedge the rewrite softened incorrectly, and added sources from my notes. I remain responsible for the final text.

If your course bans humanizers or rewriting tools, do not use them — and do not invent a disclosure for work you did not do.

Minimal (syllabus wants short)

"

Generative AI used: ChatGPT for outline options from my notes. All analysis, prose, and citations are my own. Tool output reviewed and revised by me.

No AI used (if required)

"

I did not use generative AI tools in the preparation of this assessment.

Weak vs strong wording

Cleanup tools and disclosure

Using a ChatGPT watermark remover or meaning-locked humanizer is still AI-adjacent assistance. If disclosure is required for writing tools, include it.

Honest framing:

• You cleaned paste artifacts and robotic phrasing
• You verified meaning and sources
• You did not treat the tool as undeclared authorship

PassMyEssay removes invisible ChatGPT watermarks, flags stock phrases, and humanizes with meaning locked — only when policy already allows revision support. It does not replace disclosure or policy compliance. Workflow: humanize ChatGPT text.

Pre-submit checklist

• Syllabus AI rule read for this assignment
• Tools named accurately (including PassMyEssay if used)
• Tasks listed at useful granularity
• Verification / responsibility sentence included
• Citations audited if the model suggested any
• Private prompt/draft record saved if process may be requested

Disclose like a methods note

Transparent assistance, human responsibility, no theater. Be specific. Be accurate. Ask early when the syllabus is fuzzy.

If ChatGPT touched your draft and policy allows cleanup, try PassMyEssay free — ChatGPT watermark remover plus phrase cleanup with meaning locked — then disclose the tool the same way you would name any other writing aid your course requires you to report.