AI Tells Detector

AI writing pattern checker

Paste your text. It marks the patterns people link to AI writing, says why each one stands out and shows the fix. It counts patterns and never guesses who wrote the text.

Try a sample:

By using this checker you agree to the terms.

Pasted text is checked in your browser and never sent anywhere. A web address is read once by our server and not stored.

A pattern count, not a probability. People write like this too, and AI can write without these. It cannot tell you who wrote the text.

What it looks for

29 patterns, from em dashes and “it’s not X, it’s Y” to words like delve and tapestry. Each one comes from a named source: Wikipedia’s guide to signs of AI writing, published word studies and Slate’s own copy rules. The full list, with an example and a fix for each, is on the AI writing tells page.

What it cannot tell you

Whether a person or a machine wrote something. Careful human writers trip these tells, and a chatbot can be told to avoid them. Try the Dickens sample above: a novel from 1859 still gets marked. That is why the result is a count and a list, never a score. Read more about what any AI content detector can and cannot tell you.

Use it to edit your own writing

The best use is your own draft. Find the stock phrases and rewrite them so the page sounds like you. The guide to humanizing AI text walks through each fix, and the AI text checker turns the result into a list you can tick off as you edit.

Your text stays in your browser

Pasted text is checked on your own device and never sent anywhere. A web page address is read by our server for that one request and not stored. The privacy page has the details, and the method page lists every source.

Slate Marketing LLC writes website copy for local businesses. See what Slate does.