Schools closing because of air raids. Teachers scattering across borders. Electricity cutting out mid-lesson. These are the realities facing schools across Ukraine, which entered its fourth calendar year of full-scale war with Russia in 2026.
In Ukraine, and elsewhere around the world, education systems are confronting urgent questions: How can teachers be supported when schools operate under crisis conditions? How do systems reengage students who fall behind early as a consequence and never catch up? And how do AI and technology fit into the mix?
On January 23 and 24, the Harvard Graduate School of Education held a conference to explore these questions through student-led policy analyses from a graduate class held by Fernando Reimers, the Ford foundation professor of the practice of international education. The students’ projects drew on partnerships with global ministries of education and nongovernmental organizations.
One panel—led by HGSE students Chloe Zeng, Ed.M. ’26, Lucas Kuziv M.P.A. ’26, Sicen Wan, Ed.M. ’26, and Tung Nguyen, Ed.M. ’26—focused on the use of artificial intelligence in Ukraine’s education system. Panelists discussed a central question: can AI help sustain teaching and learning when a country is at war?
Since the Russian invasion in 2022, more than 43,000 Ukrainian teachers have been internally displaced or forced to flee the country. Nearly one-third of classes are now taught fully online or in hybrid formats. Schools have been damaged or destroyed, power outages are common, and internet connectivity is unreliable, especially in frontline and rural regions.
Suddenly, emerging digital technologies (including AI) were no longer optional tools teachers weighed with an eye to integrating them fairly and effectively, but “essential infrastructure,” according to the students, for their potential to ease the teacher shortage burden.
Major global technology companies, including Google, Anthropic, and OpenAI, have stepped in to provide technical infrastructure and expertise. In 2022, Google and the United Nations Educational, Scientific, and Cultural Organization (UNESCO)—a global agency that works to promote international cooperation in education—provided 50,000 computers to help teachers in Ukraine deliver distance learning. The laptops enabled educators to continue teaching classes remotely for millions of students, many of whom had been displaced or were living in areas where schools had been damaged or destroyed.
These tech companies and international organizations have poured resources into digital education initiatives worldwide in recent years, often with an eye toward expanding access in low- and middle-income countries. Anthropic announced its “AI Literacy and Creator Collective (LCC)” to bring AI tools and training to educators in 63 countries on January 20; a day later, OpenAI announced its own “Education for Countries” initiative, to help deliver AI tools to countries around the world.
Yet, as the HGSE student team found in the case of Ukraine, investment does not automatically translate into impact.
“The problem isn’t a lack of pilot [programs] or innovation,” Zeng said. “It’s a lack of coherence.” In other words, making sure digital initiatives are implemented in a consistently effective way within a country or region.
As Zeng explained, Ukraine lacks a national framework that describes what AI-related skills teachers should develop, how progress should be measured, and how the many parallel initiatives underway should fit together. Institutions responsible for teacher training have also been weakened by staff displacement during the war. Without clear guidance and equitable distribution of resources, she said, AI adoption risks benefiting the most well-financed schools and regions while leaving others behind.
The students traced this challenge to three causes. First, there is no shared roadmap for the use of AI in education, in Ukraine or elsewhere. Second, infrastructure within the country is highly unstable, due to frequent power outages and the physical destruction of schools. Finally, the distribution of resources is uneven: teachers with greater mobility and connectivity, especially in less war-torn provinces, are far more likely to be able to take advantage of AI than those in conflict-affected areas.
To address these realities, the team compared three policy options. One would create a “voluntary micro-credential program” to train 10,000 “AI Master Teachers.” Another would establish 35 regional AI innovation hubs offering intensive, in-person coaching. And a third would launch a centrally guided national AI upskilling program, reaching teachers across the country through existing institutions. The students ultimately recommended the third approach, complemented by targeted high-intensity interventions where conditions would allow it.
Arguably, the students’ recommendation reflects a broader shift in how future education leaders are being trained at the HGSE and beyond. AI is becoming a central concern for policymakers in countries at every income level. At the conference, for instance, other HGSE master’s candidates presented projects on Uruguay’s under-financed classrooms, Thailand’s southern border provinces, and remote learning in Papua New Guinea. The education efforts in these countries provide glimpses into the future of problem-solving using AI and other digital solutions in poor or conflict-ridden nations—and the risk of framing technology as a panacea.
