Tools in this video
Transcript
Do AI detectors work? One said the US Constitution was written by AI. It was written in 1787.
OpenAI built its own detector. It caught just 26% of AI text and wrongly flagged 9% of human writing. So OpenAI shut it down.
Stanford researchers ran seven detectors on essays by non-native English speakers. 61% got flagged as AI. It was human writing.
So if a detector accuses you: a detector score is not proof.
Questions and answers
Why did OpenAI shut down its AI detector?
It caught only 26% of AI-written text and wrongly flagged 9% of human writing, so OpenAI retired it in July 2023.
Are AI detectors biased against non-native English speakers?
A Stanford study of seven detectors found 61% of essays by non-native English speakers were flagged as AI, though humans wrote them.
What should I do if an AI detector flags my work?
Remember a detector score is not proof. Ask what evidence is being used and share drafts or version history that show your process.