An AI-supervised remote exam went so badly that 58,000 students must retake it
Summary
Ars Technica reports a mass failure of an AI-monitored remote entrance exam at UNAM, revealing unusual score inflation and widespread cheating concerns. An expert commission recommended an in-person control exam affecting around 58,000 applicants, highlighting the challenges of AI-driven proctoring for large-scale assessments. The piece discusses security and policy implications of remote AI supervision in education.