Integrity built for the AI era.
AI tools make ordinary tests easy to game. iMocha locks the test environment, verifies the person and flags what a reviewer should see, across assessments, projects and Tara interviews.
Five ways people cheat. How each one is handled.
Pick a scenario to see what’s covered today and what’s coming.
Connected devices
The Safe Assessment Browser locks the test environment: blocks other software, detects virtual machines and controls what the computer can do. Every exit from the test screen is logged with its duration.
Next
Covered today.
Screen sharing
The Safe Assessment Browser blocks screen sharing and screen recording apps from running.
Next
Covered today.
Mobile phone
The camera captures the candidate throughout, so a reviewer can see phone use and repeated looking away. Unusually long completion times are flagged automatically.
Next
Object detection: phones automatically detected in the camera view.
A second person
If visible on camera, flagged today as Multiple Faces under Image Proctoring.
Next
Audio proctoring: a voice baseline at the start, with any additional voice flagged for review.
AI assistants
The Safe Assessment Browser blocks AI tools from running. Screen recording shows everything on screen. For coding tests, code replay shows whether code was typed or pasted.
Next
AI-usage detection for voice agents, plugins and generated answers.
No online test platform can see a completely separate device or hear a person out of camera view today. What we catch is the effect: the time it costs and the looking away. Object detection and audio proctoring close the remaining gaps.
Protection before, during and after the test.
Before the test
Identity verification
Government ID plus a live selfie, checked in about ten seconds. Failures are recorded.
Question randomisation
Each candidate gets a different draw from a large bank, with shuffled options.
During the test
Image proctoring
Photos every 20 to 40 seconds. Flags a missing, second or wrong face.
Video and screen recording
Continuous camera and full-screen recording, reviewable per section.
Leaving the test screen
Every exit counted and timed, with automatic termination at your limit.
Copy and paste disabled
Nothing pasted in, no questions copied out.
After the test
Colour-coded flags
Every candidate flagged in the results grid, so your team reviews the flagged few.
Every flag comes with proof
Photo, video timestamp, screen recording. Your team decides; we never auto-reject.
Completion time flags
Unusually fast or slow finishers are highlighted.
Review the flagged few, not everyone.
Results arrive colour-coded, with the evidence one click away. High scores with red flags get a second look before anyone moves forward.
One integrity layer across every format.
Skills assessments
Safe Assessment Browser, image and video proctoring, randomisation and completion-time flags.
ExploreTara interviews
Image proctoring and the Safe Assessment Browser for voice conversations.
ExploreProject-based assessments
Optional monitoring notes, AI assistant prompts recorded, never auto-failed.
ExploreStart with a focused pilot.
One or two roles, or one team. We help you pick the assessments, run them with real people and compare the results with your current process.