If you have been flagged by an AI detector for work you actually wrote, a calm, well-organized AI detection appeal letter is your single most important move. It is the document that turns a panicked "this is so unfair" into a clear, factual case that an instructor or integrity officer can act on. This guide gives you the structure, the right tone, and a full copy-paste template you can adapt in a few minutes.
A letter does not have to be long or formal to work. It has to do three things: state plainly that you wrote the work, present your evidence in order, and ask for the specific remedy you want. The reader is usually busy and has seen defensive, angry messages before. A composed letter that makes their job easy is the one that gets a fair hearing.
This page is the letter itself. For the wider picture, the hub on being falsely accused of using AI ties everything together, and the two pages you will lean on most are the deep dive on how to prove you didn't use AI (the evidence) and the walkthrough of what to do when you are accused of using AI in college (the process and your rights).
Timing matters more than people expect. Send your appeal within 24 to 48 hours of learning about the flag. Acting quickly does two things: it shows you take the matter seriously, and it gets your version on record while the work is fresh and your draft files are easy to find. Waiting a week can look like avoidance, even when it was just stress.
Who receives the letter depends on your institution's policy, so check that first. In most cases the first letter goes to the instructor who raised the flag, because the issue often resolves at that level. If your school routes academic integrity cases through a dedicated office, such as Student Affairs, an Academic Integrity Office, or a Dean of Students, follow that channel and copy the relevant person. When in doubt, address the instructor and ask, in the letter, who else should be included. Never contact the detector company (Turnitin, GPTZero, and so on); they do not adjudicate your case and have no role in the decision.
Keep the whole thing in writing. A face-to-face conversation can help, but the letter is the record that follows the case, so put your core points in the email even if you also plan to meet.
The hardest part of this letter is emotional, not technical. Being accused of cheating when you did the work is genuinely upsetting, and that feeling wants to come out as outrage or as over-explaining. Resist both. The most persuasive appeal reads like it was written by someone with nothing to hide, because you are.
A few principles keep the tone right:
If you are too rattled to hit that tone, draft the angry version first, get it out of your system, then rewrite. The reader only sees the second draft.
Every effective AI detection appeal letter moves through the same five beats. You can write each in two or three sentences.
The subject line should tell the reader exactly what this is before they open it. Use a format like:
Appeal of AI Detection Flag - [Your Name] - [Course Code/Name]
This is specific, searchable, and professional. Avoid vague lines like "Question about my grade" or emotional ones like "This is a mistake." The goal is for the right person to find and prioritize your message instantly.
Open by naming the situation directly: the assignment, the date, and the fact that an AI detector returned a result you are contesting. Then state, in one sentence, that you wrote the work yourself. No throat-clearing, no apology for something you did not do.
This is the heart of the letter. List your proof in order of strength, briefly, and reference the files you are attaching. The single most persuasive item is your writing process: version history from Google Docs or tracked changes in Microsoft Word, which shows the document being built over time. We cover the full evidence pack in how to prove you didn't use AI, so keep the letter itself to a tidy summary and let the attachments carry the detail.
Add a short, factual paragraph explaining that a detector result is not proof. This is where attributed facts do real work. A Stanford study (Liang et al., 2023) found that several GPT detectors flagged essays by non-native English speakers as AI-generated at an average false-positive rate of 61.3 percent, far above the rate for native writers. Independent estimates of general false-positive rates run roughly 5 to 20 percent. And Turnitin itself states that its AI writing score should not be used as the sole basis for an adverse action against a student. You are not claiming detectors are useless; you are noting, accurately, that they are probabilistic and known to misfire, which is exactly why a human review is warranted.
Close by stating clearly what you want: a human review of your evidence, or an oral defense (a viva) where you discuss your sources and reasoning in your own words. Make it explicit that you are asking for a decision based on the full picture rather than a single tool's number. A concrete, reasonable request is far easier to grant than a vague plea to "please reconsider."
Your attachments are the part that actually wins the case, so name them in the letter and include them with the email. The core set:
A short, labeled list of attachments at the end of the letter ("Attached: 1. Google Docs version history; 2. Outline; 3. Second detector report") makes the package easy to follow.
Adapt the bracketed parts, delete anything that does not apply, and keep it concise. This fits a single screen, which is exactly what you want.
Subject: Appeal of AI Detection Flag - [Your Full Name] - [Course Code/Name]
Dear [Professor/Instructor Name or Academic Integrity Office],
I am writing regarding [Assignment Name], submitted on [Date] for
[Course Code/Name], which was flagged by an AI detection tool as
possibly AI-generated. I want to address this directly: I wrote this
assignment myself, and I would welcome the chance to demonstrate that.
I understand why the flag has to be taken seriously, and I am not
disputing that the tool returned a result. I am asking that the result
be reviewed in light of the evidence below, which shows my writing
process from start to finish.
Evidence of my work (attached):
1. Version history from [Google Docs / Microsoft Word tracked changes],
showing the document being written and revised over [time period].
2. My [outline / earlier drafts / research notes], which trace how the
ideas developed.
3. A second, independent AI detector report on the same text, for
comparison with the original score.
I would also note, respectfully, that AI detectors are known to produce
false positives. A 2023 Stanford study found these tools flagged
non-native English writing as AI-generated at an average rate of about
61 percent, and Turnitin states that its AI score should not be the sole
basis for action against a student. I share this not to dismiss the
concern but to explain why I believe a human review is the fair way
forward.
I would be grateful for one of the following: a review of my evidence by
a person, or an oral discussion where I can walk through my sources,
argument, and choices in my own words. I am happy to provide anything
further that would help.
Thank you for your time and for considering this fairly.
Sincerely,
[Your Full Name]
[Student ID, if applicable]
[Contact information]
If your case is escalating to a formal hearing rather than a quick email exchange, the guide to being accused of using AI in college covers the procedural side, your rights, and how the timeline usually unfolds.
A good letter is partly about what you leave out.
TextSight is an AI-detection and writing-trust tool. We help you check and document your own work; we do not help anyone disguise AI writing or evade an institution's review. No detector can prove who wrote a piece of text, and ours is probabilistic too, so use a second score as one supporting data point in your evidence pack, never as proof on its own. The strongest defense is always your writing process.
TextSight's free tier gives you daily scans with sentence-level highlights, so you can see which lines carry the AI signal and why. Every score is a probability, not proof, and we say so plainly.