Skip to content

Guide · 5 min read

Flagged as AI, but you wrote it.

A detector score is not proof. Here is what it means, and what helps when your own work is questioned.

By TemplateBreakerPublished

Short answer

Stay calm and gather your process evidence: drafts, version history, notes and sources. A detector score is an estimate, not proof, and detector makers themselves warn against relying on a score alone. Ask to talk through how you wrote the piece, and read your school's or employer's policy on detector results.

What does an AI detector score actually mean?

An AI detector score is a model's estimate that a text resembles machine writing. It is not a record of how the text was made. Detectors look at signals such as how predictable the wording is and how evenly the sentences run, and plenty of human writing shares those signals.

Even the companies that build these tools say so. Turnitin's own guidance says the percentage on its AI writing indicator should not be used as the sole basis for action or as a definitive grading measure. OpenAI withdrew its own AI text classifier in July 2023, citing its low rate of accuracy.

What should I do first?

Do not panic, and do not delete anything. Collect everything that shows how the piece came together before you reply. Then ask, politely and in writing, what the concern is based on and whether you can talk it through.

What evidence helps show I wrote it?

Version history is the strongest record most writers have. Google Docs keeps it under File, Version history, and Word keeps it when the file lives in OneDrive. Save outlines, early drafts and notes, not just the final file.

Your research trail helps too: bookmarks, highlighted readings, citations and search history show where your ideas came from. If you wrote by hand first, photos of those pages count.

How do I talk to my instructor or editor about it?

Offer to walk through how you planned, researched and revised the piece. Explaining your own argument, and why you made particular choices, is often more convincing than any document. Keep the tone factual; the goal is a fair review, not a debate about software.

Does my school's policy say how detector results can be used?

Read it. Some institutions limit how detector results may be used, and some have turned the tools off. Vanderbilt University, for example, disabled Turnitin's AI detector in 2023, citing concerns about false positives and about bias against non-native English writers.

Why do non-native English writers get flagged more often?

A 2023 study by Liang and colleagues, published in the journal Patterns, found that widely used GPT detectors misclassified essays by non-native English writers as AI-generated much more often than essays by native speakers. Simpler vocabulary and more regular sentence structure can look predictable to a detector, even when a person wrote every word.

Can I see which parts of my writing read as formulaic?

Yes. Our free check highlights stock openers, stacked transitions and even rhythm in text you paste, with no account, so you can see what a reader or a detector might react to. It explains patterns; it does not prove who wrote the text, and it should not be used to argue a score either way.

See alsoCheck your own writing freeSigns of AI writingHow AI detectors work

Questions, answered

Can a teacher prove I used AI with a detector?

No detector proves authorship, and detector makers themselves warn against relying on a score alone. A score can start a conversation; it should not end one.

What evidence shows I wrote it myself?

Version history in Google Docs or Word, saved outlines and early drafts, notes, highlighted readings, citations and search history. Being able to explain your argument and choices in person also helps.

Why would my own writing be flagged?

Detectors look for predictable wording and even rhythm, which careful, formal or heavily edited human writing can have. Research by Liang et al. (2023) found detectors often misclassified essays by non-native English writers as AI-written.

Should I rewrite my essay so it is not flagged again?

Do not rewrite work to chase a score. If you want to understand which phrasing stands out, the free check at /check explains formulaic patterns in your own text. It does not prove authorship, and no edit can promise a detector result.

Sources and further reading