A useful originality check shows both a likelihood and the reasoning, and lets you tune how strict it is.
Artificial Intelligence Detector combines a sensitivity selector with a full signal breakdown for a more transparent estimate.
How to use Artificial Intelligence Detector
- Set the sensitivity to lenient, balanced or strict.
- Keep the detailed signal breakdown on for full transparency.
- Inspect each signal, then judge the passage yourself.
Use cases
- Reviewing a passage with both a score and its signals.
- Tuning strictness while keeping the breakdown visible.
- Teaching what statistical patterns AI detectors rely on.
Good to know
Artificial Intelligence Detector reports an AI-likelihood estimate alongside the four signals that produced it, and lets you set how strict the labelling is. It measures patterns, not authorship, so it is an indicator and not a verdict: always confirm with manual review. The heuristic is English-only, needs about 40 or more words, and analyzes the first 40,000 characters of very long input with a visible note.
Frequently asked questions
Is an AI-likelihood estimate the same as proof?
No. It estimates how closely text matches statistical patterns common in AI writing; it cannot prove who or what wrote a passage.
How do the sensitivity and detail options work together?
Sensitivity sets how strict the label is, while the detail breakdown shows the signals behind the score, so you can see why a passage was labelled.
Does it support non-English text?
No. It is English-only and will flag non-English input as unsupported rather than guess a score.
Should I act on the score alone?
No. Use it as a starting point and confirm with a human read before making any decision.