HomeResources › How to Write an AI Detection Appeal Letter

How to Write an AI Detection Appeal Letter

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.

Try the AI detector free How our detector works

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).

When to send it, and to whom

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 tone that works: calm, not defensive

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:

  • Acknowledge the flag without conceding it. Open by recognizing that a detector returned an AI result. You are not arguing the score never happened; you are explaining that it is wrong about you.
  • State the truth once, plainly. "I wrote this assignment myself" is stronger than three paragraphs insisting on it. Confidence reads as honesty.
  • Stay factual about the tool. You can note that AI detectors are documented as unreliable without attacking your instructor or sounding like you copied a Reddit thread. Cite, do not rant.
  • Ask, do not demand. You are requesting a fair review, not issuing ultimatums. Cooperative language gets cooperative responses.

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.

The structure of a strong appeal letter

Every effective AI detection appeal letter moves through the same five beats. You can write each in two or three sentences.

1. A clear subject line

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.

2. Acknowledge the flag and state your position

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.

3. Present your evidence systematically

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.

4. Note the documented unreliability of AI detectors

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.

5. Ask for a specific remedy

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."

What to attach

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:

  • Version history or tracked changes. Export or screenshot your Google Docs version history (File, then Version history, then See version history) or save the Word document with tracked changes on. This timeline of edits is the strongest single piece of evidence, because it shows the work forming over hours and days.
  • Drafts, outlines, and notes. Any earlier rough version, your outline, brainstorming notes, or annotated readings. They corroborate the timeline.
  • A second independent detector score. Run the same text through a different detector and include the result. The honest way to frame this: if a second tool returns a low or different score, the variance between two probabilistic tools is itself meaningful, and a low score is one data point that supports your case. It does not prove you wrote the work, because no detector can prove that, but combined with your drafts it strengthens the picture. You can get a free second-opinion score and sentence-level breakdown from the TextSight AI Detector to attach alongside the original report.
  • Research trail. Browser history, library checkout records, or saved sources that show how you gathered material.

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.

The copy-paste template

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 few things not to do

A good letter is partly about what you leave out.

  • Do not threaten or invoke lawyers in a first letter. That escalates a problem that often resolves with a calm exchange. Keep legal options in reserve.
  • Do not fabricate or backfill evidence. If you did not keep drafts, say so honestly and lean on what you do have, such as an oral defense. Inventing a paper trail is the one thing that turns a defensible case into a real violation.
  • Do not over-attack the detector. One factual paragraph on reliability is persuasive. Five paragraphs reads as deflection.
  • Do not send it angry. Sleep on it if you can, or have a calm friend read it first. Tone is doing as much work here as content.
FAQ

Frequently asked.

How long should an AI detection appeal letter be?
Short. Aim for under one page in the body of an email, with the detail living in your attachments. A busy reader is more likely to act on a tight, clear letter than a long, emotional one. The template above is a good length to copy and adapt.
What is the most important thing to include?
Your writing process. Version history from Google Docs or tracked changes in Word is the strongest single piece of evidence, because it shows the document being built over time in a way a detector cannot. Everything else supports that.
Can a second AI detector score prove I am innocent?
No, and it is important to be honest about that. No detector can prove a negative, and a second tool is also probabilistic. What it can do is add a useful data point: if two tools disagree, that variance shows the score is not reliable on its own, which supports your request for a human review. Treat it as one item in the evidence pack, not the whole case.
Who should I send the appeal to?
Follow your institution's policy. Usually the first letter goes to the instructor who raised the flag, and from there it may move to an academic integrity office or Dean of Students. Address the instructor, and ask in the letter who else should be copied if you are unsure.
How quickly do I need to respond?
Within 24 to 48 hours of learning about the flag, if you can. Acting promptly looks responsible and keeps your draft files easy to locate while everything is fresh.
What if I do not have any drafts saved?
Be honest about it rather than inventing something. Lean on whatever you do have, your research history, your ability to discuss the topic in depth, and offer an oral defense (a viva). Being able to explain your own argument in person is itself strong evidence that you wrote it.

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.

Check your own writing. Free, no signup.

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.

Start free, no card See the methodology