AI can fake the answer, the face, and the voice. Lotan verifies that the person who
earned the result is the person who sat the session — continuously, from first frame
to last, with an evidence trail a registrar or a regulator can actually read.
Continuous matchFlagged for human reviewIdle / between items
How it works
Everyone else verifies a moment. We verify the whole session.
One engine, three guarantees. An ID snapshot at the door tells you who walked in —
not who stayed, and not who did the work.
01
Real human
Liveness checks resilient to injected video and face-swap, bound to an identity at
the start of the session.
The anchor matches the stakes: campus SSO for enrolled students, government ID for
licensure and hiring. Universities never hand us a document per exam.
02
Same human
Re-verification runs for the entire session, not once at the door. Substitution
mid-session is the attack nobody else is looking for.
Every checkpoint is a tick on the strip above — the record shows continuity, not a
single hopeful frame.
03
Unassisted
Screen, environment and behavioural signals surface coaching, second devices and
live model assistance — the 2026 attack, not the 2019 one.
Signals are evidence for a reviewer, never an automatic verdict against a candidate.
Where it runs
The engine doesn't care what the session is called.
An exam, a certification, an interview — all are sessions where somebody has to prove
they can do the thing. Same infrastructure, different stakes.
Certification & licensure
High-stakes remote delivery without a test centre. Per-exam pricing, evidence
packages built for appeals and audit.
Education
Installs inside the LMS over LTI — instructors launch a verified assessment from
Canvas without a second login or a new workflow.
Interview integrity
The same continuous verification pointed at hiring, where a proxy candidate costs
far more than a failed exam.
Our stance
Integrity without surveillance theatre.
First-generation proctoring earned its backlash: room scans held unconstitutional at a
public university, biometric class actions, students treated as suspects by default.
We took the opposite side of every one of those decisions — because a tool that
institutions are embarrassed to defend is a tool they eventually drop.
No room scans. Ever. We do not ask anyone to film their home.
A human makes every integrity decision. The system flags; it never accuses.
Written consent, published retention schedule, no sale of biometric data — built to the strictest state standard, not the weakest.
Accessibility is a design input. Assistive tech, atypical movement and poor bandwidth are not integrity signals.
Tenant isolation is absolute. Your sessions are never visible on our staff surfaces.
Get started
AI made competence easy to fake.
We make it provable. If you issue a credential that has to mean something —
a certification, a course, a hire — we should talk.