The candidate on the call might not be the person you hire.
SyncLens screens a recorded interview and tells you whether the face, the voice, or both were synthetic, and which part of the call the signal is in.
The scenario
A recruiter runs a clean remote interview. The recording looks normal. Weeks later, the person who starts on day one does not match the person who answered the questions.
Synthetic candidates in recorded interviews are a live problem in 2026. A recruiter watching in real time is not equipped to catch a good one.
The problem, in hiring terms
You are not trying to prove fraud. You are trying to decide whether to trust a recording enough to move a candidate forward, and to have something you can point to if that decision is questioned later.
A yes-or-no detector does not give you that. A per-segment report with visible evidence does.
Which result matters most here
Fake video, real audio is the one to watch. A real voice answering real questions, over a face that is not the speaker's, is the shape of an interview where a stand-in feeds answers to an on-camera actor.
Fake video, fake audio matters too. That is what a fully synthetic feed looks like. Real video, fake audio is uncommon in this setting.
- Read first
- Fake video, real audio
- Also relevant
- Fake video, fake audio
Reading the report
The report is built to be read by the person making the call, not only by a specialist.
1. Result and confidence. Plain language first. If confidence is low, that is stated before the result.
2. The timeline. See whether the whole interview is in question or only a stretch of it. A break that lines up with the hard technical questions is worth a second look.
3. Tagged frames. A non-technical reviewer can see what the model reacted to, without reading a spectrogram.
4. Provenance. Model version and file hash, so the report can be reproduced if the hire is later challenged.
Fits the hiring workflow
Send the recorded interview from your applicant tracking system or your video platform to the API when processing finishes. Get a report back before the debrief.
Batch upload covers a full hiring round in one pass.
Read the API reference