As artificial intelligence (AI) takes on increasingly critical roles — from managing power grids to piloting autonomous vehicles — making sure these systems are safe has never been more important. But how can we be certain that the AI controlling them can be trusted?  

A research team at the University of Waterloo is addressing this question using tools from applied mathematics and machine learning to rigorously check and verify the safety of AI-driven systems.  

The first national randomized trial of public Montessori preschool students showed stronger long-term outcomes by kindergarten, including elevated reading, memory, and executive function as compared to non-Montessori preschoolers. The research also appears highly actionable for policymakers, because the results found the Montessori programs delivered better outcomes at sharply lower costs. The study of 588 children across two dozen programs nationwide shows an imperative to follow and study these outcomes through graduation and beyond.