At a glance
- The claim: a detector score proves a named author used AI.
- The fact: independent research has documented false positives and unequal error rates.
- The missing evidence: direct, reliable provenance information connecting the author to the alleged process.
- The person making a public accusation is accountable for checking it before causing reputational harm.
The claim: the detector caught them
Paste prose into a detector, receive a percentage, post the screenshot, declare the author guilty. Are we serious? The percentage is the output of another automated model. It is not testimony, a document history, an admission, or a record of how the text was created.
Reality check: the tools produce false accusations
A peer-reviewed Stanford-led study tested seven widely used detectors on human-written essays. The tools frequently misclassified writing by non-native English speakers as AI-generated; the researchers reported that more than half of the TOEFL essays were falsely flagged, while essays from U.S.-born eighth graders were handled far more accurately. They also showed that prompting changes could help generated text evade detection.
That is not proof that every detector always fails. It is proof that the output depends on the tool, sample, language patterns, model family, editing, and decision threshold. The number is a model’s estimate under particular conditions. It is not a chain of custody.
The author bears the damage; the accuser owns the claim
Writers who use formal prose, predictable sentence structures, editing tools, translation assistance, or English as an additional language may be disproportionately exposed to suspicion. Publicly naming a person based on a detector score can cause reputational harm that no later correction fully reverses.
What would count as better evidence
If provenance matters, seek direct evidence: disclosed workflows, document history, credible reporting, contractual records, or platform-required declarations. Detector output can be one investigative signal, but it should be interpreted cautiously and never presented as self-proving evidence. Authors remain accountable for honest claims about their work. Accusers are equally accountable for distinguishing suspicion from fact.
Sources and documents
These links support the factual claims in this explainer. Interpretive conclusions are identified in the article and should be read with the underlying record.