Table of Contents
Most interview processes depend on one assumption: the person in the first round is the same person in the second, the third, and eventually the job. Seat swapping is what happens when that assumption breaks.
A candidate interviews for the rounds they can pass and brings in someone else for the rounds they cannot. The real applicant might handle the recruiter screen and behavioral interview, while someone more qualified takes the technical round. In some cases, two people split the process live, switching who is on camera depending on the question. This kind of seat swapping leaves no obvious trace.
Same name. Same application. Different person in the seat.
And because everyone involved is a real person, it often does not look fake.
Why seat swapping stays invisible
Each interview looks normal by itself. A real person joins the call, answers the questions, and may even pass an ID check. If you only look at one round at a time, there is often nothing obvious to investigate.
The fraud only becomes visible when you compare the rounds against each other. That is the comparison most hiring processes are not built to make.
Remote hiring makes this easier to miss. Interview loops now often include several rounds, handled by different people on different days. A recruiter runs the screen, an engineer runs the technical, a hiring manager runs the final, and everyone debriefs on skills, experience, and fit.
However, identity continuity is rarely part of the process. No one is usually responsible for confirming that the person in round three is the same person from round one. So the seat swap slips through the handoffs. By the time the mismatch becomes obvious, the offer is often already out. The person starts the job, struggles, and gets written off as a bad hire.
Sometimes that is true. But sometimes the simpler explanation is that the person who passed the interview was never the person who showed up to do the work.
How you catch seat swapping
The usual checks do not solve this problem.
ID verification, background checks, and deepfake scans are useful, but they usually confirm a single moment. Seat swapping happens across moments. It is not always about whether one call looks suspicious. It is about whether the person stays consistent from one round to the next.
To catch it, you have to connect the rounds.
That means comparing the signals that should stay relatively stable in a real candidate: their face, voice, cadence, and the way they explain their thinking. If the person in the final round no longer matches the person from the technical, or their skill level suddenly jumps beyond anything the earlier rounds suggested, that is the signal.
No single mismatch proves fraud because every candidate can have a good or bad call and cameras, rooms, lighting, and internet connections can change. The pattern is what matters. A seat swap shows up when a candidate stops being consistent in ways that nerves or bad wifi cannot explain.
And that is hard to catch by hand. Recruiters are not going to rewatch every earlier round before the final. Hiring managers are not going to line up recordings and compare them in real time. Done manually, it would slow hiring down too much.
Detection tools can do that work in the background. They can compare each round against the others and flag meaningful changes without adding another step to the process.
How Tofu catches seat swapping
This is what we built Tofu to do.
Tofu tracks identity across the full interview loop, not just one moment in it. It compares signals from the first screen to the final round and flags when the person in the seat appears to change.
It works across Zoom, Google Meet, Teams, and other video interview platforms without adding extra work for recruiters. The goal is simple: catch the seat swap before it turns into a bad hire.
If your interview process has more than one round, you already have the gap. If you want to see what's slipping through yours, come talk to us.
