Measuring AI impact on learning outcomes

AI is transforming learning, but evidence gaps remain. This event explores what works, for whom, and how to measure AI’s impact at scale to inform smarter investments for children and adolescents.

Pedro Irala Tapari is 8 years old, a member of the Ava Guaraní indigenous community, and attends third grade at Arroyo Verde School, located in a rural area of ​​Santa Rosa del Aguaray, San Pedro Department, Paraguay. Thanks to the Let's Go to School! project
UNICEF/UNI872831/Brizuela

About

The #GLOCAL2026 session is on: 2 June 2026 | 10 AM Panama/ 11 AM New York

 

Artificial Intelligence (AI) is entering classrooms with potential to transform personalized learning, tutoring, and teaching. Recent studies highlight AI's rapid adoption in parts of Latin America and the Caribbean, and rigorous evidence on its impact shows promising results. However, many other AI applications across the region have not integrated strong impact-focused evaluative measures needed for scale-up for real improvements for children and adolescents. Evidence gaps persist where AI studies often lack long-term outcome data or fail to measure unintended effects, including around bias and exclusion of marginalized groups. While applications globally demonstrate promise, policymakers still face uncertainty around what works, for whom, and under what conditions in order to invest in new technology for scale-up. This event will bring together leading government, research, and evaluation experts to discuss promising results of AI applications in the learning sector, assess what’s known, and identify gaps for evaluating AI’s contribution to outcomes at scale for children and adolescents. The event aims to inform evidence agendas for measuring AI impacts on learning outcomes.

 

Register here.

Author(s)
Fiorella Haim, María Paz Monge, Juliette Norrmen-Smith, Martín Elías De Simone, Michael Craft