Mexico’s UNAM conducted its first fully remote entrance exam using AI‑driven webcam monitoring, but the results showed a five‑fold surge in top scores, forcing 58,000 candidates to retake the test. The fiasco raises serious concerns about AI‑based exam security.

Key Takeaways

  • Nearly 160,000 applicants took the remote exam
  • Top scores increased 5× (16.3% vs 3.5%)
  • 58,000 students must retake the exam

UNAM, Mexico’s largest university, launched its first completely remote entrance exam this summer, requiring candidates to use a lockdown browser and AI‑powered webcam proctoring software over several weeks.

The results were shocking. Between 2021 and 2025, only 3.5% of test‑takers scored 100 + out of 120 questions. In 2024 that share exploded to 16.3%, a five‑fold jump that calls the exam’s validity into question.

Historical Background

Founded in 1551, UNAM has traditionally relied on in‑person written examinations. The 2024 shift to a digital, AI‑monitored format was intended to broaden access, but technical glitches and inadequate oversight turned the experiment into a disaster.

Why This Matters

BozokMedia analysis shows that deploying AI‑based proctoring without robust safeguards can undermine academic integrity, jeopardizing future digital assessments.

"AI proctoring compromised exam integrity," says Dr. Ana Martinez, academic assessment expert.
Year% Scoring 100+
2021‑2023 Avg.3.5%
202416.3%
Did You Know?: UNAM is the largest university in Latin America and was founded in 1551.

Frequently Asked Questions

Q1: Why did the AI proctoring tools fail?
A: Software recognition errors, network latency, and poor student compliance were primary factors.

Q2: What is UNAM doing for the next exam?
A: The university is adding human monitors, updating the software, and considering a hybrid testing model.