UNAM orders 58,000 retakes after AI-proctored exam
Mexico’s largest university will rerun a remote entrance exam after anomalous score results.
Why it matters
The case highlights operational and trust risks when institutions use AI-enabled proctoring for high-stakes assessments. It also shows that remote exam systems can create downstream fairness and validity problems at large scale.
The key points
- 1.UNAM will require 58,000 applicants to retake the entrance exam.
- 2.The remote exam used a lockdown browser and AI webcam proctoring.
- 3.Top scores rose sharply compared with 2021-2025 results.
UNAM, Mexico’s largest university, ordered 58,000 applicants to retake its entrance exam after its first fully remote administration using a lockdown browser and AI-powered webcam proctoring produced anomalous results. Nearly 160,000 applicants took the test from late May through early June, and 16.3% scored 100 or more out of 120, compared with 3.5% between 2021 and 2025.
⚡ Try this today
Audit score distributions and escalation plans before using AI proctoring in high-stakes exams.
Sources & original reporting
This brief summarizes and links to reporting from the publishers below.
Enjoyed this brief? Get the next one in your inbox.
More in Policy & Safety
Woman alleges stepfather used Grok on childhood photo
A TechCrunch report says a woman claims Grok was used to turn a childhood image into explicit imagery.
Anthropic details how Claude text watermarks will work
The company says Claude will use SynthID-Text and plans a detection API.
Judge flags hidden AI prompts in court filing
A Connecticut judge said the prompts did not affect the records-access case but set a dangerous precedent.