Lucas and Luna explore how one startup, ClassDoze, uses AI to detect student disengagement during live lessons. The system analyzes webcam eye gaze, posture, and micro-expressions to flag students who are zoning out, bored, or frustrated. Lucas shares a case study from a pilot in three high schools where disengagement alerts helped teachers re-engage students mid-class, increasing reported attentiveness by 15%. Luna questions privacy concerns and the risk of surveillance culture. The hosts discuss the trade-off between personalized attention and constant monitoring, and what this means for the future of classroom dynamics. They also touch on the current regulatory environment for AI in education, and how schools are balancing innovation with student well-being. This episode is part of the Fexingo Business podcast network.