Why have some universities banned Proctorio and similar AI proctoring tools?
TL;DR: Mostly privacy, bias, and trust — not proctoring itself. Students at schools like UBC and Miami University pushed back over invasive data collection and biased facial detection, and one vendor even filed a lawsuit against a critic that backfired publicly. The lesson for institutions isn't "don't proctor" — it's "proctor transparently, with evidence, not a black-box score."
The Evidence
In September 2020, Proctorio sued Ian Linkletter, a learning-technology specialist at the University of British Columbia, for sharing unlisted YouTube links to the company's own training videos in a critical thread about the software. The Electronic Frontier Foundation called it a SLAPP suit — a lawsuit designed to silence a critic rather than win on the merits. Inside Higher Ed and the EFF covered the case as it dragged on for years; Proctorio eventually dropped it. The episode became a rallying point for student and faculty distrust of proctoring vendors generally, not just Proctorio.
Separately, students at multiple universities — including a widely circulated petition at University of Illinois Urbana-Champaign — objected to browser-lockdown and webcam tools collecting data they felt was disproportionate to a course exam: room scans, keystroke patterns, gaze tracking, all funneled into an opaque "suspicion" score neither the student nor sometimes the instructor could fully see the reasoning behind. Combined with the facial-detection bias findings covered in academic literature (see: does facial recognition proctoring show racial bias), the pattern that drove bans wasn't proctoring itself — it was proctoring without transparency or appeal.
What Got Banned vs. What Institutions Still Use
| Practice that drew bans | What institutions replaced it with |
|---|---|
| Opaque "suspicion score" with no evidence shown | Trust reports showing the actual evidence behind every flag |
| Facial detection as sole violation trigger | Facial detection as one signal among several, human-reviewed |
| No student-facing appeal process | Documented, exportable evidence a student can contest |
| Vendor legal threats against critics | Vendors that publish their methodology and accuracy data |
How to Avoid Repeating the Same Mistakes
- Choose a vendor that shows evidence — screenshots, timestamps, violation timelines — not just a numeric score.
- Require human review before any flag becomes an academic-integrity case.
- Publish your proctoring policy to students before the exam, including what's monitored and why.
- Give students a documented way to see and contest evidence.
- Ask your vendor how facial-detection accuracy varies across skin tones — and get the answer in writing.
FAQ
Is AI proctoring generally being phased out? No — demand has grown as AI-assisted cheating has grown. What's changed is the standard: vendors and institutions can no longer get away with a black-box "risk score" and no evidence trail.
Did the Linkletter lawsuit involve cheating detection accuracy? No, it was about him sharing publicly available training-video links. It became a flashpoint for broader distrust, not evidence the software didn't work.
What does Neuroxa do differently? Every session ends in a defensible trust report: violation timeline, evidence snapshots, and a one-click PDF export — so a flag is never just a number a student has to take on faith.
Should small institutions worry about legal risk from proctoring vendors? Ask any vendor directly about past disputes with users or critics, and review the data-retention and privacy terms before signing.
By Pinal Dave Last updated: 2026-07-25