AI in Global Education: Expand Human Agency, Do Not Outsource It

White Paper Executive Brief | Trevor Mishler | August 2026

AI is entering education systems that are profoundly unequal in connectivity, teacher capacity, language access, infrastructure, and institutional readiness. The technology can amplify good pedagogy and capable educators, but it can also amplify weak content, surveillance, bias, and unequal access.

The evidence suggests that AI can improve immediate performance, but stronger output is not automatically durable learning. The most promising approaches align tools to clear educational goals, preserve productive struggle, keep teachers in the learning relationship, and treat equity as a design constraint.

High-consequence uses such as grading, placement, admissions, cheating detection, or risk prediction require stronger evidence, transparency, human review, and meaningful appeal. Scale should follow evidence rather than novelty or adoption metrics alone.

Bottom line: use AI to expand human agency, not to outsource it.