UNICEF event to examine AI learning outcomes

A gLocal Evaluation Week 2026 session will examine how to measure the impact of AI on learning outcomes for children and adolescents.

The event, titled ‘Measuring AI impact on learning outcomes’, is scheduled for 2 June and will focus on evidence gaps around AI use in education, particularly in Latin America and the Caribbean.

The session will discuss how AI is entering classrooms through personalised learning, tutoring, and teaching tools. UNICEF says some applications show promising results, but many AI tools used across the region lack rigorous impact-focused evaluation measures needed to assess whether they improve learning outcomes and can be scaled effectively.

The discussion will bring together government, research, and evaluation experts to assess existing evidence, identify promising results, and examine gaps in measuring AI’s contribution to learning outcomes at scale.

Participants will also consider unintended effects, including bias and the exclusion of marginalised groups. UNICEF says policymakers still face uncertainty over what works, for whom, and under what conditions when deciding whether to invest in AI tools for education.

Speakers listed for the session include Fiorella Haim of Ceibal, Martín Elías De Simone of the World Bank, Juliette Norrmen-Smith of UNICEF’s Office of Innovation, María Paz Monge of J-PAL-LAC, and Michael Craft of UNICEF. The event will include simultaneous Spanish-English interpretation.

Why does it matter?

The session highlights a key challenge in AI and education: adoption is moving faster than evidence. As AI tools enter classrooms, policymakers need stronger evaluation methods to determine whether they improve learning outcomes, work for different groups of children, and risk reinforcing bias or exclusion.

Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!