This research report from the Brookings Institution’s Center for Universal Education shares findings from a yearlong global study, which argue that the risks of generative AI in education outweigh its advantages. After conducting interviews, focus groups, and consultations with over 500 education stakeholders (students, teachers, parents, education leaders, and technologists) and completing a comprehensive review of over 400 studies, the report shares potential benefits and risks of AI in education. The report is organized into six sections: (1) introduction of the study, (2) context of children’s AI use, (3) potential benefits, (4) potential risks, (5) weighing risks and benefits, and (6) recommendations for future AI use. It calls on government agencies, education funders, technology companies, school districts, organizations, researchers, and families to take action in creating a more responsible and ethical future for AI use. The report concludes with actions rooted in three pillars (Prosper, Prepare, Protect) to minimize the risk of AI harm in schools and maximize its potential to support learning.