Interview integrity & fraud detection
Fake and AI-assisted candidates are 2026's #1 hiring threat — deepfake interview fraud grew 1,300% in two years. Gaugely verifies who you're really talking to: optional government-ID + liveness verification, on-device presence checks, tab-switch and plagiarism flags, and an append-only integrity log on every scorecard. Flags go to a human — nothing here rejects a candidate.
- →Optional identity verification (government ID + liveness selfie) before the interview starts — checked on our servers, not in the candidate's browser
- →On-device presence signals: no-face / multiple-faces flagged for review
- →Tab-switch, plagiarism, and prompt-injection attempts logged as integrity events
- →Repeated tab-switching can end an interview early at a limit you set — the candidate is told up front and can retake
- →Every flag lands on the scorecard with evidence — a human decides what it means
Gaugely closes the loop: screen CVs → interview the best (voice or video) → compare scorecards → re-engage your past top candidates. No other tool feeds your own interview signal back into screening.
Questions
Can AI detect fake or proxy candidates in interviews?
Gaugely layers signals instead of guessing: verified government ID + liveness before the call, presence analysis during it, and tab-switch/plagiarism/prompt-injection flags on the scorecard. No signal ever rejects a candidate — a human reviews the integrity log and decides.
Can a candidate suppress the tab-switch and presence flags?
A determined one can: those signals are reported by the candidate's own browser, so they are corroboration rather than proof, and a clean integrity log is not evidence of an honest interview. What cannot be suppressed is the identity check — government ID plus a liveness selfie, verified server-side before the call. Enable it on the roles where fraud actually matters and treat the browser signals as what they are: a reason to look closer.
Does identity verification hurt the candidate experience?
It adds about a minute, is disclosed on the consent screen, and is optional per job — most teams enable it only for sensitive or remote-critical roles. Verified candidates get a "Verified identity" badge, which protects honest applicants from being out-competed by fraud.
How do you identify a fake candidate in a video interview?
Look for signals rather than a single tell: an identity that was never verified against a government ID, a face that disappears or multiplies mid-answer, answers that arrive after a tab switch, and phrasing that matches a model's output more than a person's. Any one of these can be innocent — a cluster of them is worth a human review.
What is interview fraud?
Interview fraud is when the person assessed is not the person who would do the job — a proxy sitting the interview, a deepfaked face or voice, or an undisclosed AI feeding answers in real time. It has moved from rare to routine in remote hiring, which is why verification now happens before the interview rather than at offer stage.
Can you detect deepfakes in an interview?
No detector is perfect, so Gaugely does not sell one. It raises the cost of faking instead: liveness-checked ID before the call, presence analysis during it, and a full integrity log a human reads. The goal is evidence a recruiter can act on, not an automated fraud verdict.