Many institutions secure online exams with a lockdown browser. It is a sensible first step: while the exam is open, switching apps, opening other sites and copying text all become harder. But a browser can only control what happens inside the browser. It cannot see the desk, the room, a phone, a second person or software running outside its reach, and most attempts to get around a locked exam do not attack the browser at all. They go around it.
This white paper sets out a layered model for online exam security: a locked exam session, identity verification that continues through the exam, AI-powered threat detection, optional desktop-level checks on Windows, and trained people who review what the technology flags before any decision is made. It explains which layer addresses which threat, from second devices and extra monitors to virtual machines, remote access, overlays, AI tools and impersonation.
It also covers why an AI flag is a reason to look and never a finding, what a defensible case record should hold (event timeline, clips, reviewer observations, audit trail and evidence pack), six recommendations for a fair exam policy, and an honest table of what each proctoring setup can and cannot see.
The paper describes threats in general terms only and contains no instructions for bypassing exam software.
Who it is for
Controllers of Examinations and exam office teams, academic integrity and malpractice committees, IT and information security heads, and academic leaders deciding how much proctoring each exam needs.
What you will learn
- What a lockdown browser can and cannot do on its own.
- The six common online exam threats, and why most of them work around the browser.
- The five layers of modern exam security, with a threat-by-layer map.
- Why human-in-the-loop review matters, from flag to committee to appeal.
- What a case record needs so a decision holds up on appeal.
- Six steps for a fair and secure online examination policy.


