Guide

Are AI detectors biased against non-native English writers?

The evidence says yes, and the mechanism is well understood. Detectors reward linguistic unpredictability — which is exactly what writing in a second language tends to have less of.

Last updated July 26, 2026 · 8 min read

Key takeaways

  • A 2023 Stanford study found seven GPT detectors misclassified more than half of TOEFL essays by non-native English writers as AI-generated, while classifying native-speaker essays almost perfectly.
  • The bias is structural: detectors score linguistic predictability, and second-language writing is typically more regular in vocabulary and sentence construction.
  • In the same study, rewriting the essays with richer vocabulary sharply reduced false flags — evidence that the detectors were reading proficiency, not authorship.
  • This lands hardest on international students, who often face the highest stakes and the least institutional standing to contest an accusation.
  • Improving your English is a legitimate response. Concealing authorship is not — and the distinction matters both ethically and practically.

What the research found

The central study is Liang, Yuksekgonul, Mao, Wu and Zou, 'GPT detectors are biased against non-native English writers', published in the Cell Press journal Patterns in July 2023.

The design was simple. The researchers took TOEFL essays written by non-native English speakers and essays written by US eighth-graders — native speakers, and younger writers at that — and ran both sets through seven widely used GPT detectors.

Essay setAuthorOutcome
US eighth-grade essaysNative English speakersClassified as human-written almost perfectly
TOEFL essaysNon-native English speakersMore than half misclassified as AI-generated
TOEFL essays, vocabulary enrichedSame non-native writersMisclassification dropped sharply
Summary of the Liang et al. (2023) finding across seven detectors.

That third row is the one that settles the question. The authorship never changed. The only variable was how linguistically varied the text was. When the researchers prompted a model to rewrite the same essays with more sophisticated language, the detectors largely stopped flagging them.

A detector whose output changes that much based on vocabulary richness, while authorship stays constant, is not measuring authorship.

Why this study carries weight

It is peer-reviewed, it tested multiple commercial detectors rather than one, and it included a manipulation that isolated the mechanism. The comparison group being younger native speakers makes the result harder to explain away as a writing-quality effect.

Why second-language writing trips the signal

Detectors are built around two measurable properties, and second-language writing tends to score low on both for reasons that have nothing to do with machines.

A narrower working vocabulary

Writing in a second language usually means drawing on a smaller set of reliably known words. You choose the word you are confident is correct rather than the unusual one that might be slightly off.

That is good judgement. It also means the next word in your sentence is more often the statistically likely one — which is the definition of low perplexity, the primary thing detectors measure.

More consistent sentence construction

Second-language writers frequently rely on sentence patterns they have been explicitly taught and know to be correct. The result is even, well-formed, structurally consistent prose.

Detectors read that evenness as low burstiness. Native speakers, writing intuitively, tend to produce more erratic rhythm — fragments, asides, sentences that run long because the thought did.

The proofreading trap

There is a second-order problem that specifically affects ESL writers. If you are less confident in your English, you are more likely to run your work through a grammar checker or translation tool before submitting it.

Those tools move text toward conventional, expected phrasing — lowering perplexity further. So the writer most likely to be flagged is also the writer most motivated to use the tools that increase the chance of being flagged.

Worth being clear about

Using a grammar checker on your own writing is not academic misconduct at any institution we are aware of. But it can move your text toward the statistical profile a detector treats as machine-like, and it is better to know that in advance.

Who carries the cost

The consequences fall unevenly, and they fall hardest on people least able to absorb them.

  1. International students. Often paying substantially higher fees, sometimes on visas contingent on good academic standing, and frequently least familiar with the institution's appeals process. An accusation carries more risk and is harder to contest.
  2. Researchers publishing in English. Most academic publishing happens in English regardless of the author's first language. Detector screening at journals introduces a bias that has nothing to do with the quality of the research.
  3. Professionals writing for global employers. Reports, proposals and internal documents increasingly pass through AI-detection tooling. Being flagged in a workplace has no appeals process at all.
  4. Job applicants. Some hiring platforms screen cover letters for AI use. A non-native applicant writing carefully in English can be filtered out before a human reads a word.

Institutions have started responding

Some universities have stepped back from automated detection over reliability concerns — Vanderbilt disabled Turnitin's AI detector in August 2023 and published its reasoning. Turnitin's own guidance frames its score as the start of a conversation rather than a determination of misconduct. If you are contesting a flag, your institution's own policy language is often the most useful document you have.

What you can actually do about it

There is a legitimate response and an illegitimate one, and the line is worth stating plainly before the practical advice.

Improving the range and rhythm of English that you wrote is ordinary editing — the same thing a copy editor or a writing centre would help you do. Passing off work you did not do is misconduct regardless of what tooling is involved. Everything below assumes the first case.

Practical steps:

  1. Protect the record by default. Write in a tool that keeps revision history — Google Docs, Word, Notion. Do not draft elsewhere and paste in a finished block. A visible history of incremental work is the strongest evidence of authorship you can have, and it costs nothing to create.
  2. Deliberately vary sentence length. Follow two long sentences with a short one. This raises burstiness, which is the property detectors measure — and it genuinely improves readability for any reader.
  3. Add specifics only you would know. A detail from a source you actually read, an example from your own context, a particular number. Generic writing is what reads as machine-generated; specificity is both better writing and a stronger signal of a real author.
  4. Keep some of your own phrasing. Resist the urge to smooth every sentence into the most conventional possible form. Slight unconventionality is a human signal. Correct is not the same as maximally expected.
  5. Check before you submit, not after. Knowing how your text scores lets you decide what to do while you still have options. After an accusation, your leverage is much lower.
  6. If accused, argue process rather than score. Draft history, notes, sources and a willingness to discuss the content are far more persuasive than a competing detector reading. Our guide on false positives covers the full sequence.

Where a tool like ours fits — and where it does not

We build an English rewriting tool, so treat this section as interested rather than neutral. It is worth being specific about what it does.

Our humanizer takes English text you paste and rewrites it with more varied vocabulary and sentence rhythm while keeping your meaning. It returns several options, and you choose the one that still sounds like you. There are no tone or strength settings — you paste text and get alternatives back.

For ESL writers, the honest framing is that this is an editing aid for English you already wrote. It is the same category of help as a writing centre appointment or a native-speaker friend reading your draft. It cannot verify authorship, and no tool should claim otherwise.

Two real limits

The model is optimized for English; results in other languages may be inaccurate, so this is not a translation workflow. And detector scores shift as detectors are retrained — anyone promising a permanent guarantee against a specific detector is overselling.

0 / 300 words

Paste English you wrote to see the alternatives. Free plan covers 500 words per month, 300 per request.

Sources

Frequently Asked Questions

Are AI detectors really biased against non-native English speakers?

Yes, according to peer-reviewed research. A 2023 Stanford study published in Patterns tested seven GPT detectors on TOEFL essays by non-native English writers and essays by US eighth-graders. The detectors classified the native-speaker essays almost perfectly but misclassified more than half of the non-native essays as AI-generated.

Why does my English writing get flagged as AI?

Because detectors measure linguistic predictability rather than authorship. Writing in a second language typically uses a narrower working vocabulary and more consistent sentence patterns, which produces low perplexity and low burstiness — the two properties detectors treat as machine-like. Your writing is being read as predictable, not as artificial.

Does using a grammar checker make it worse?

It can. Grammar and translation tools move text toward conventional phrasing, which lowers perplexity further. Using them on your own work is not misconduct, but it does shift your text toward the profile detectors flag — which is a particular problem for ESL writers, who are the most likely to use them.

What is the evidence that detectors measure proficiency rather than authorship?

In the Stanford study, the researchers rewrote the same non-native essays with richer vocabulary and re-ran the detectors. Misclassification dropped sharply even though the authorship had not changed. A tool whose verdict changes with vocabulary while authorship stays constant is not detecting authorship.

Is it ethical for an ESL writer to use a rewriting tool?

Editing English that you wrote is ordinary practice — it is what a writing centre or a copy editor does. Submitting work you did not produce is misconduct regardless of the tooling. The distinction is whether the ideas and the drafting are yours. If your institution requires disclosure of AI assistance, follow that policy.

Can I prove I wrote something myself?

Process evidence is the strongest thing available. Preserve your document's revision history before copying the text anywhere, keep drafts, notes and sources, and be ready to discuss the content. Incremental development over time is hard to fabricate and much more persuasive than a competing detector score.

Do universities know about this problem?

Increasingly, yes. Vanderbilt disabled Turnitin's AI detector in August 2023 and published its reasoning, and Turnitin's own guidance describes its indicator as a starting point for discussion rather than a determination of misconduct. If you are contesting a flag, check what your institution's policy says about detector output as sole evidence.

Polish English you wrote yourself

Paste your draft and get alternatives with more natural range and rhythm, with your meaning intact. You pick the version that still sounds like you.

Try it free

Free plan includes 500 words per month. English only. No credit card required.