Cite as: Real Problem AI problem “Why can't a 200-person company tell if their final-round candidate is a deepfake proxy?”. Opportunity score 8.5 out of 10 (severity 9, AI feasibility 8, market signal 9, competition gap 7). Category HR Tech & Recruiting. Trend Vision. Source signal: CXOToday + TechRadar + Sherlock AI on AI interview fraud. Canonical URL: https://www.realproblem.ai/archive/why-can-t-a-200-person-company-tell-if-their-final-round-candidate-is-a-deepfake-proxy.
Why can't a 200-person company tell if their final-round candidate is a deepfake proxy?
50% of companies have hit AI-driven deepfake interview fraud; HireVue / Greenhouse don't detect it; 59% of hiring managers suspect candidates are using AI to misrepresent themselves.
Who has it: Recruiters and hiring managers at 50-2000 person companies running remote video interviews on Zoom / Teams / Meet in 2026.
- Severity9/10
- Feasible today8/10
- Market signal9/10
- Competition gap7/10
What people report
The source reports that many hiring managers suspect candidates of using AI to misrepresent themselves, and that live detection often means switching interview platforms.
Our summary of the public post linked below, not a quote. Nobody submitted it to Real Problem AI.
CXOToday + TechRadar + Sherlock AI on AI interview fraudWhat nothing does yet
A bot-as-attendee that joins any Zoom / Teams / Meet call, runs deepfake + identity-mismatch detection passively, and pushes a one-line risk score into Greenhouse / Ashby, no platform switch.
Tools that solve this
4 named so far, and where each one stops.
- Sherlock AI
97% detection claim but requires platform switch.
- Pindrop
Voice-biometric focus; weak on synthetic video.
- Reality Defender
Enterprise-priced; not Greenhouse-native.
- HireVue
Adds detection inside its own platform; ATS lock-in.
Why it is archived
Retired in the 2026-10-25 freshness audit: deepfake candidate screening from May 2026; identity verification vendors (Persona, CLEAR, Socure) and ATS add-ons now sell this, and newer live entries cover AI hiring.
How you might build it
- 01WebRTC + Zoom / Teams / Meet bot SDK
- 02Vision + audio synthetic-media detection
- 03Greenhouse / Ashby webhook
- 04Auditable risk-score log
Founders working in this area
- Ben Colman, Co-founder and CEO, Reality Defender
Not affiliated with Real Problem AI.
#H13 · Canonical URL: https://www.realproblem.ai/archive/why-can-t-a-200-person-company-tell-if-their-final-round-candidate-is-a-deepfake-proxy