What it is
Zero-trust exam security is the principle that the candidate's device is untrusted by default. The assessment is protected by controlling the network path and verifying device posture, rather than by trusting that the machine is clean.
Why it matters
Legacy proctoring assumes a benign candidate operating a benign device; every modern cheating tool weaponizes that assumption, which is why a zero-trust posture is a prerequisite for defensible high-stakes assessment.
How Aiseptor implements it
Aiseptor applies zero-trust principles from enterprise network security to remote assessment: every device joins an ephemeral enclave, every destination is authorized explicitly, and the integrity of the session is continuously re-verified.
Definition
Canonical definition
Zero-trust exam security is an architectural doctrine that treats the candidate's device as untrusted for the full duration of an assessment. Instead of building defenses around assumptions about the candidate's environment, such as that the installed applications are innocuous, that the camera catches every cue, or that the browser is the only relevant attack surface, the model asserts that the device may be fully compromised and designs controls around that worst case. The exam platform therefore enforces a minimal, explicit set of allowed destinations; verifies device posture before and during the session; and treats any deviation as a signal rather than a proof-of-cheating. This mirrors the shift enterprise security underwent over the last decade: the perimeter-and-trusted-interior model (a corporate firewall, an assumed-safe internal network) gave way to zero-trust because attackers routinely operated from inside the perimeter. Exam integrity is undergoing the same shift, a decade later, for the same underlying reason: assuming the candidate's device is the trusted interior has stopped being a safe assumption. Zero-trust exam security is the category Aiseptor defines; it is the application, to assessment integrity, of the same principles that have reshaped enterprise network architecture over the last decade.
In practice
A concrete example
A university IT department historically trusted any device on the exam-hall Wi-Fi because access to the hall itself was controlled at the door. When exams moved remote, that same trust model was ported over by default: if a candidate could log into the exam portal, the device behind that login was implicitly trusted. A zero-trust redesign removes that inheritance entirely. Logging in proves who the candidate is; it proves nothing about what else is running on their machine. The device only earns trust for the destinations its session policy explicitly grants, re-verified continuously, not once at login and then assumed for the next two hours.
Citations
- [1]Aiseptor architecture whitepaper (public version) (2026)
- [2]NIST Special Publication 800-207, Zero Trust Architecture (2020)