This gives a rough, heuristic "AI-style signal" score based on surface patterns like sentence-length uniformity, vocabulary variety, and common AI phrasing. It is not a reliable or accurate AI detector, and both human and AI writing can score anywhere on the scale.
Gives a rough, heuristic signal based on writing-style patterns — sentence uniformity, vocabulary variety and common AI phrasing. It is not a reliable detector and misclassifies plenty of text in both directions. Please don't use it to accuse anyone, or as the basis for a grading, disciplinary or publishing decision.
Rough heuristic only — always treat the result as a loose signal, never as proof.