Detection · Virtual cameras & deepfakes
Detect virtual cameras and deepfakes in remote interviews.
OBS, ManyCam, and injected video feeds let a candidate replace the live webcam with pre-recorded video, a manipulated stream, or a deepfake — so the person on camera may not be the person interviewing. Capifiq surfaces a substituted camera feed and seals it as evidence.
The threat
The face on camera may not be live.
A virtual camera sits between the real webcam and the meeting, and can feed the interview anything: a pre-recorded loop, an altered stream, a stand-in, or a real-time deepfake.
Combined with AI-generated resumes and synthetic voices, it is how a candidate profile can be partly — or entirely — fabricated. Gartner projects one in four candidate profiles will be fake by 2028.
Tools in this category
- OBS Virtual Camera — free, ubiquitous
- ManyCam — virtual camera effects
- Injected and virtual camera devices
- Deepfake and face-swap tools
- Pre-recorded video loops
- …and new tools shipped constantly
The blind spot
Why the interviewer accepts the feed.
Modern virtual cameras produce a smooth, believable image, and the manipulation happens before the video ever reaches the meeting.
A convincing feed
The output is smooth and believable, and interviewers naturally accept the video they are shown as live.
Before the meeting sees it
The substitution happens ahead of Zoom or Teams, so the meeting app just receives a normal-looking feed.
Legitimate, and everywhere
Streamers and creators use these tools daily, so their existence isn't proof — and a denylist can't tell.
What Capifiq surfaces
- Virtual camera devices and active video sources
- A change to the camera source during the session
- Coverage that isn't limited to named tools
- Cryptographically hashed, sealed evidence for review
The coverage
Caught by the source, not the picture.
Capifiq surfaces indicators that the camera feed is being substituted or routed through a virtual device — not by judging the image, but by verifying the camera source.
The finding is timestamped and sealed for the interviewer to review.
The full picture
On-device is only half of it.
Capifiq detects any unauthorized assistance running on the candidate's device — this is one of the common vectors, not a fixed list. And for help that never touches the device, interviewing methodology deters what no software can see.
Hidden AI assistants
Desktop and web tools that feed answers on a hidden overlay.
Remote-control sessions
Another person viewing or driving the candidate's machine.
Virtual machines
A second environment concealing unauthorized tools.
Browser-based assistants
Extensions and web apps that generate answers.
Display tampering
Overlays and extra displays hiding content.
Off-device cheating
Deterred with interviewing methodology — the second pillar.
Questions
Common questions.
Can Capifiq tell if the video is a deepfake?
Does using OBS mean someone is cheating?
What about synthetic voice or AI resumes?
What about a brand-new virtual-camera tool?
Get started
Make sure the person on camera is real.
Run Capifiq alongside the Zoom, Teams, or Meet calls you already do, and see what it surfaces.