Occurred: May 2021
Page published: December 2021 | Page last updated: December 2025
Dartmouth College’s Geisel School of Medicine falsely accused students of cheating based on misinterpreted "pings" from automated learning management software, leading to a major due-process scandal.
Dartmouth College's Geisel School of Medicine launched a "dragnet" investigation into dozens of medical students for alleged cheating on remote, closed-book exams.
The school’s IT department retroactively audited a year’s worth of log data from Canvas, the school's learning management system, to identify instances where students ostensibly accessed course materials during active exam windows.
By March 2021, the school formally charged 17 students with honour code violations. The potential harms were severe: students faced expulsion, suspension, or permanent "unprofessional conduct" marks on their transcripts - disciplinary actions that can effectively end a medical career by disqualifying students from future residencies.
The investigation caused significant mental health strain, with students reporting isolation, weight loss, and suicidal ideation.
Some students were ordered to repeat an entire academic year, which would have cost approximately USD 70,000 in additional tuition, plus living expenses. Others felt forced to hire specialised legal counsel to fight the school’s "black box" evidence, incurring significant out-of-pocket costs.
Independent technical analysis later proved that the "activity" was actually automated background syncing - essentially, devices "pinging" the server for updates without any human interaction, even if it was happening without the student knowing about it, on a device that was not in use during the exam.
The school failed to disclose that it was secretly monitoring background activity. When challenged, administrators were slow to admit that their "proprietary" detection method could not distinguish between a manual page view and an automated system refresh.
Dartmouth eventually dropped all charges against the students, acknowledging that the technical data that formed the basis of the charges was insufficient. The Dean of the Geisel School, Duane Compton, apologised to the students and the entire student body.
For the impacted: For the accused, the event was a traumatic betrayal of trust. While all charges were eventually dropped and an apology issued in June 2021, the reputational and psychological damage was deep.
For society and education: The scandal serves as a landmark warning against "algorithmic authoritarianism" in education. It highlights the dangers of relying on black box data to make life-altering disciplinary decisions.
For future policy: The incident accelerated the shift away from invasive remote proctoring. Dartmouth subsequently moved toward in-person exams and explored open-book formats, acknowledging that surveillance technology often creates a "panopticon" effect that prioritises policing over actual learning.
Black box
In science, computing, and engineering, a black box is a system which can be viewed in terms of its inputs and outputs (or transfer characteristics), without any knowledge of its internal workings.
Source: Wikipedia
Developer: Canvas
Country: USA
Sector: Education
Purpose: Detect and prevent cheating
Technology: Learning management system
Issue: Accountability; Accuracy/reliability; Fairness; Transparency
https://www.vnews.com/Geisel-investigates-potential-cheating-during-exams-39856089
https://apnews.com/article/medical-schools-fa16b8e4f1c8a63787070c739f48613c
https://www.nytimes.com/2021/05/09/technology/dartmouth-geisel-medical-cheating.html
https://thecollegepost.com/dartmouth-med-students-cheating-probe/
AIAAIC Repository ID: AIAAIC0619