Visual media
Art, UI screenshots, stylized portraits, social compression and unseen generators can shift model scores.
False-positive handling
A signal can be wrong.
Design for correction.
False positives are expected model failures, not edge cases to hide. Review source context before any consequential decision.
Open the free checkerScope before score.
Every result is a warning signal. The evidence and limitations below define what this route can and cannot support.
Text
Formulaic human writing, editing, language and short samples can overlap with learned AI-style patterns.
Audio
Noise, music, mixed content, codec changes and speaker/domain shift can invalidate a speech score or force unknown.
Do not infer more than the lane measured.
- Do not treat a raw score as a calibrated probability.
- Provide correction and appeal paths wherever people can be affected.
- Prefer unknown over forcing unsupported content into human or AI.