What should I do if an AI detector flags my writing?

Don't panic. Gather evidence of your writing process — Google Docs version history, draft files, research notes, and timestamps. Write a calm, factual response citing your evidence and referencing the detection tool's own guidance that it is not standalone proof of misconduct.

Why would an AI detector flag writing that I wrote myself?

AI detectors analyse statistical properties of text output — vocabulary patterns, sentence structure, perplexity. Formal academic writing, carefully edited prose, and writing by non-native English speakers share these properties with AI-generated text. The detector cannot determine how the text was produced from the text alone.

Is an AI detector result proof of misconduct?

No. AI detector vendors themselves — including Turnitin and GPTZero — explicitly state that their tools are aids to educator judgment and should not be used as sole proof of academic misconduct. A score is a starting point for investigation, not a finding.

Scripli

Your writing was flagged as AI — but you know you wrote it.

Here's how to prove it,
step by step.

AI detectors don't actually detect AI use — they estimate whether writing resembles AI output. Formal, clear, or precise writing gets flagged constantly. You likely already have strong evidence. Let's find it.

What do I say? What evidence do I have?

Start here

Gather your evidence — version history, drafts, research notes, timestamps.

Take screenshots or export records so you have something to attach.

Send a clear, evidence-based response. Use the template below.

You likely already have everything you need — even if you haven't looked for it yet.

Email template

What to say to your professor or editor

Be direct, professional, and evidence-first. Don't apologise for something you didn't do. Copy and adapt this template:

Email template — adapt to your situationCopy & paste

Subject: Response to AI Detection Flag — [Your Name], [Assignment Name]

Dear [Professor / Instructor Name],

I'm writing to respond to the concern raised about my [essay/assignment] "[Title]" submitted on [Date].

I wrote this work myself. I understand AI detectors can produce false positives, particularly for writing that is formal, well-structured, or follows academic conventions closely — and I want to provide evidence that demonstrates my writing process.

Evidence I can provide:

1. Google Docs version history showing my document evolving from a blank page over [X] writing sessions between [start date] and [submission date].

2. [Any other evidence: research notes, browser history, drafts, tutor emails, etc.]

I would welcome the opportunity to walk through this evidence with you, answer any questions about the work, or complete a brief oral examination on the content to further demonstrate my understanding.

I take academic integrity seriously, which is why I want to address this directly and transparently. Please let me know how you would like to proceed.

Sincerely,
[Your Name]
[Student ID / Course]
[Contact details]

Key principles: Lead with evidence, not emotion. Request a meeting rather than demanding a decision. Offer an oral exam — this is your strongest counter because AI doesn't understand what it writes. Stay professional and factual throughout. Academic integrity panels respond much better to documented evidence than to emotional appeals.

Open in email app

Opens your email app with the template pre-filled. Edit the brackets before sending.

Honest about what Scripli can and can't do

Scripli proves authorship going forward — for writing sessions recorded through Scripli. For work already submitted, we can't retroactively generate a certificate. What we can do: help you understand the evidence you already have, and ensure every piece of writing you produce from now on comes with permanent, verifiable proof.

Evidence checklist

What evidence do you already have?

Check each item. You likely have more than you realise.

Google Docs version history

Very strong evidence

If you wrote in Google Docs, your entire edit history is preserved — every keystroke, pause, and revision. This is powerful evidence of a human writing process.

How to get it: How to access: File → Version history → See version history (or Ctrl+Alt+Shift+H)

Draft versions and autosaves

Strong evidence

Word, Google Docs, and most editors autosave regularly. Multiple saved drafts showing evolution of the document demonstrate a human revision process.

How to get it: Check OneDrive (Word), Google Drive version history, or your Downloads folder for earlier exports.

Browser and device timestamps

Strong evidence

Your computer logs when files were created and modified. Screenshots of file metadata (Date Modified, Date Created) show you were actively editing over time.

How to get it: Right-click any local file → Properties (Windows) or Get Info (Mac) to see timestamps.

Research notes and search history

Supporting evidence

Your research trail — browser history, saved tabs, notes — shows the intellectual work that preceded the writing. AI doesn't leave a search history.

How to get it: Export browser history or screenshot your notes app. Even rough notes in a physical notebook are evidence.

Email and messages about the work

Supporting evidence

Emails to a tutor asking for feedback, messages to a classmate discussing the topic, screenshots of research conversations — all show you engaged with the material.

How to get it: Search your email for the document title, assignment name, or the date you were working on it.

If you have version history or drafts, you're in a strong position.

Most false flags are resolved when clear evidence of the writing process is provided. The combination of a timestamped edit history and a calm, evidence-first response is typically enough to close the case.

Step-by-step guide

How to export Google Docs version history

Version history is your strongest evidence if you wrote in Google Docs. It shows every edit, with timestamps, going back to the moment you started typing.

1
Open your Google Doc.
2
Click File in the top menu.
3
Click Version history → See version history.
4
A sidebar opens showing every save point with timestamps.
5
You can click any timestamp to see exactly what the document looked like at that moment.
6
Take screenshots of the earliest versions and the full timeline — these show your document evolving from a blank page.
7
For a stronger export: click the three dots next to a version and select "Restore this version" to see the edit-by-edit view. Screenshot this.
8
(Optional) Right-click the sidebar → Print / Save as PDF to create a portable record.

If you didn't use Google Docs: Microsoft Word autosaves to OneDrive and keeps version history in the Version History panel (File → Info → Version History). Notion, Notion Pages, and most modern editors also keep edit histories. Even a text file with multiple renamed saves (essay-draft1.docx, essay-v2.docx) is evidence.

Get a Writing History Report from your Google Doc

New

Connect your Google Doc and we'll generate a verifiable record of when you started writing, how many sessions there were, and how the document evolved — based on Google Drive's revision timestamps.

This creates a Writing History Report — not a Human Authorship Certificate. It shows when writing began and how it evolved, but does not capture keystroke behavior.

Scripli reads only revision timestamps and character counts — never your document text. You can revoke access at any time in your Google account settings.

Why are you being flagged if you wrote it yourself?

AI detectors don't actually detect AI use. They estimate whether writing resembles AI output — which means clear, structured, or formal writing is often flagged, even when written entirely by a human.

Research consistently shows false positive rates of 20–50% on human writing. ESL writers, autistic writers, and anyone trained to write formally (academic writing courses, law students, professional writers) are disproportionately affected. The detector isn't broken — it's just answering the wrong question. "Resembles AI output" is not the same as "was produced by AI."

The strongest counter to a detector result is not an argument — it is evidence. Version history, revision patterns, research trail. Things that AI cannot produce because AI does not have a process.

This situation is exactly what Scripli is designed to prevent.

For every piece of writing you create going forward, Scripli records your writing process and generates a permanent, verifiable proof record. So if you're ever questioned again, you won't need to gather evidence — you'll already have it.

Works in Google Docs, Word Online, Canvas, Substack, and more via Chrome extension
Your document text never leaves your device — only proof metadata is recorded
Free for every writer — no credit card, no institution required
Each session produces a shareable verification link anyone can check

Protect your future writing — free For students

No school or institution required • Free tier available • Public verification for verifiers