How do students coordinate cheating rings using group chat apps during online exams?
TL;DR: By opening a group chat — GroupMe, Discord, WhatsApp — on a phone or second device outside the camera's view, then posting exam questions for others to answer collectively. It's collusion, not individual cheating, and it's why mass academic-misconduct cases involving dozens or hundreds of students keep happening.
The Evidence
This isn't a hypothetical. In 2020, Georgia Tech investigated a large group of students — reported in the Atlanta Journal-Constitution and covered by the New York Post and IBTimes — for suspected collusion during at-home online physics exams, warning students to self-report or risk failing grades. Boston University ran a parallel investigation the same semester. The pattern in these cases was consistent: students used group chat apps to share exam questions and pooled answers in real time while individually logged into separate, unproctored exam sessions.
The mechanism is simple because it doesn't require any technical bypass — no VPN, no virtual machine, no lockdown-browser exploit. It just requires a second device outside the frame of a single webcam, which is exactly why single-signal webcam monitoring misses it: the collusion happens off-screen, and one exam session in isolation looks clean.
Individual Cheating vs. Coordinated Rings
| Individual cheating | Coordinated ring | |
|---|---|---|
| Detected by | Single-session webcam/screen review | Cross-session pattern analysis |
| Common tell | Eyes off-screen, phone visible | Similar wrong answers across students, submission timing clusters |
| Tools used | Notes, a second tab | Group chat apps on a second device |
| Why it's harder to catch | It usually isn't | Requires comparing sessions, not just watching one |
How to Catch and Prevent It
- Monitor screen and browser activity, not just webcam feed — collusion often correlates with tab-switching or unusual pacing even when the camera looks clean.
- Randomize question order and pull from a larger item bank so answers can't be called out and reused verbatim.
- Set tight per-question or per-section timers to shrink the window for real-time answer-sharing.
- Review submission-timing and answer-pattern clusters across the whole cohort, not session by session.
- Use a proctoring platform with a defensible evidence trail so a mass-misconduct case — if it happens — holds up in a hearing.
FAQ
Does webcam monitoring alone catch group-chat cheating? Not reliably. The chat happens on a second device out of frame. Screen and browser-behavior monitoring plus cross-session pattern review catch what a single webcam misses.
How common are these mass cheating rings? Documented cases exist at multiple large universities, typically surfacing when answer patterns across a big cohort are statistically implausible to be independent.
What's the first sign an instructor should look for? Clusters of identical wrong answers or near-identical timing across students who shouldn't have coordinated.
Can question randomization alone stop this? It raises the bar significantly but doesn't eliminate it — pair randomization with monitoring and timing analysis for real protection.
By Pinal Dave Last updated: 2026-07-25