Education is one of AI’s most exciting frontiers, but we’re still figuring out how to measure its actual effect on learning. Most current research just looks at test scores, which misses the bigger picture of how students use AI day-to-day and how that shapes their progress over time. To fix this, a team from OpenAI, Estonia’s University of Tartu, and Stanford’s SCALE Initiative created the Learning Outcomes Measurement Suite—a framework for tracking learning outcomes longitudinally across different settings. It’s being validated through a randomized controlled trial, with partners like Arizona State University, UCL, and MIT Media Lab involved. The goal is to eventually make this suite a free public resource for schools worldwide. Why it matters: This isn’t just another study—it’s a desperately needed standard for understanding if AI tools actually help students learn deeper, not just perform better on tests.