Respond promptly, calmly, and in writing. State that you wrote the work, ask for the exact concern and applicable policy, and assemble evidence showing how the work developed. Do not confess to conduct that did not occur, and do not alter the submitted file while the matter is being reviewed.
1. Ask for the specific allegation
“This was flagged by AI” is not enough information to answer. Ask which detector or review method was used, which passages raised concern, whether the score is the only evidence, and what institutional process applies. Keep the message neutral. Your goal is to understand the claim and create an accurate written record.
Thank you for letting me know. I wrote this assignment and would like to respond with evidence of my process. Could you please share the passages or report that raised concern, the relevant course policy, and the next step for review?
2. Build an authorship evidence packet
Use original files or exports that preserve timestamps. A useful packet can include:
- document version history or tracked changes;
- an outline, handwritten notes, mind map, or research questions;
- earlier drafts, including incomplete or weaker versions;
- annotated sources, library records, citation-manager entries, or reading notes;
- feedback from a teacher, tutor, or peer and the revision made afterward;
- a short timeline connecting research, drafting, feedback, and submission.
Do not manufacture evidence or backdate a document. If you drafted in a plain text editor without history, say so and rely on the evidence you do have: sources, notes, knowledge of the argument, file metadata, and your ability to explain the work.
3. Explain the essay, not the detector
Prepare to explain why you chose the thesis, how each source supports it, what you changed between drafts, and which parts were difficult. Mark two or three passages you can discuss in detail. This is usually more useful than submitting screenshots from other detectors with lower percentages.
Detector disagreement can show that scores are unstable, but it still does not prove authorship. Research on AI detector false positives recommends caution when a result is used alone. Focus your response on affirmative evidence that the work is yours.
4. Disclose allowed tools accurately
If you used spellcheck, a citation manager, translation assistance, speech-to-text, or a grammar tool, describe exactly what it did. Check the assignment rules before labeling a tool “allowed.” A local rule-based tool such as our Non-AI Grammar Checker does not generate prose or upload text, but the course policy still controls what may be used.
5. Request a fair human review
Ask the reviewer to consider the full packet, your explanation, and any evidence beyond the detection score. The BSSS academic-integrity guide is one institutional example that calls for investigation, a fair hearing, and an opportunity to provide evidence of authorship. Procedures differ, so use your school's published policy and deadlines.
6. Escalate through the stated process
If the first review does not resolve the issue, ask for the decision and reasons in writing. Note the appeal deadline, permitted evidence, and who receives the appeal. A student adviser, ombuds office, union, counselor, or trusted faculty member may help you understand the process. This guide is practical information, not legal advice.
A concise response template
I am writing to confirm that I authored the submitted work. I have attached my version history, outline, earlier drafts, source notes, and a timeline of revisions. I can explain the thesis, sources, and development of the flagged passages. Because an AI score is not direct evidence of authorship, I respectfully request that the work and this supporting evidence receive human review under the published academic-integrity process.
Before the meeting
- Keep the submitted file and evidence unchanged.
- Bring the assignment prompt and AI-use policy.
- Write down dates, names, and decisions after each conversation.
- Meet every response or appeal deadline.
- Read why an essay can be flagged as AI and the research summary on false positives.