What should I do before submitting work checked by ChatGPT?

Do not rely on ChatGPT to prove or disprove AI authorship, in either direction. If you need a signal, use a detector that publishes its method and limitations. Keep your drafts and notes so you can show how your work developed.

Does ChatGPT detect AI?

AI chatbot (OpenAI). Last updated 2026-06-23. Every claim here links to a source at the bottom.

How does ChatGPT check for AI?

ChatGPT is a language model, not a detector. When asked if text is AI-written it produces a plausible-sounding answer based on patterns in its training, not a calibrated probability. It has no special insight into whether it or another model wrote a passage, and it cannot recognise its own past output.

What this means for you

If a teacher or editor pastes your writing into ChatGPT and asks 'is this AI?', the answer carries no evidential weight. It can flag genuine human writing as AI and clear actual AI text just as easily. A purpose-built detector with a published method is a more honest signal, and even that is not proof.

Limitations to know

OpenAI withdrew its own AI Text Classifier in July 2023 over low accuracy.
ChatGPT gives different answers to the same text on different runs.
It cannot reliably identify its own previous output.

What you should do

Do not rely on ChatGPT to prove or disprove AI authorship, in either direction.
If you need a signal, use a detector that publishes its method and limitations.
Keep your drafts and notes so you can show how your work developed.

Check your own writing first

Is It AI? is free, with no signup. It highlights the passages that look AI-generated and explains why, so you can revise in your own voice before you submit.

Try Is It AI? free