Deepfake Detection: How to Spot AI Video

'Deepfake' technically refers to media where a generative model has swapped, synthesized, or altered a person's face or voice in video — but the term is now used loosely for any AI-generated or AI-manipulated footage. Both matter for the same reason: the footage shows something that didn't happen the way it appears to.

Unnatural blinking patterns (too little, or too regular) is one of the older, still-sometimes-useful tells. Look at the edges where a face meets hair, ears, or the background for blurring or flickering, and check whether lighting on the face actually matches the lighting in the rest of the scene. For audio-paired video, mismatched lip sync — especially on consonants — is often the most noticeable giveaway.

Automated video detectors look for statistical signs that footage was generated or altered, sometimes including artifacts that only show up across motion. The AI Video Detector on this site returns one AI-generation probability with ML, OCR and metadata sub-signals. It is not a dedicated face-swap detector and returns no separate face-manipulation score for video.

Check the source — who posted it first, and does their account have a track record. Look for the same footage from an independent source. An automated detector can add one more signal. And treat the result the same way as any other detector output: a confidence signal to act on carefully, not an unappealable verdict — especially before sharing or reporting on a clip further.