About the Report
Developed jointly by UNICEF and EdTech Hub, this report explores the relationships between digital and AI-enabled technologies and learning, teaching, equity, safety and system sustainability.
It brings together global evidence reviews, think pieces, and country case studies from researchers, policymakers, practitioners, and technical experts working across the digital education systems. Through their evidence-informed perspectives, the contributors map the evolving landscape of digital education and show how digital and AI-enabled tools are shaping the many purposes and goals of education.
Its five thematic chapters examine the best available evidence, surface points of tension, and set out recommendations for a more equitable future that goes “beyond digital as usual”. It is written for the governments, implementers, donors, researchers and education leaders making these decisions now.
Why the Sector Needs This
After a quarter of a century of promising claims about technology’s transformative potential, the global learning crisis remains acute. 273 million children are out of school, and a large proportion of those in school are unable to read. For millions more, learning is delayed or disrupted – including by gender inequality, disability-related exclusion, conflict- and climate-related school closures, and persistent financing gaps.
Digital and AI-enabled technologies could accelerate progress for the children falling furthest behind, or deepen the harm already being done. The direction is not yet set. What is already visible is what happens when it is left unshaped: wasted budgets, locked-in contracts and widening gaps. This report offers the evidence, frameworks and recommendations the sector needs to shape that direction.
Fragmented Practice
Digital and AI tools are entering classrooms faster than the sector can agree on how to use them well. Without shared norms and standards, practice varies widely between systems, making it harder to learn what works and to hold anyone accountable for outcomes.
Wasted Resources
Twenty years after One Laptop per Child, the sector risks betting once more that new technology offers a simple solution to a complex problem. Money spent on devices, licences and pilots that were never going to shift learning is money taken from the approaches that would have. Without an honest look at what has already been tried, interventions will absorb time and money that the sector cannot afford to waste.
Vendor Dependence
When governments adopt technologies without clear accountability frameworks, they risk long-term reliance on vendors whose incentives may not align with the public good. Strong public governance is needed to keep learning outcomes, not market interests, at the centre of decision-making.
A Widening Divide
Without deliberate intervention, a twin-track system will continue to emerge: low-cost, unverified, addictive-by-design AI for learners in low-income contexts, and premium, human-supported provision for those in high-income countries. The result is a widening gap in which the children who most need strong provision are the ones least likely to get it.
Now is the time to raise the bar for how we use technology in education, including AI. Business as usual in education is not enough. Business as usual in digital education is not enough.”
– Pia Rebello Britto, Global Director Education, UNICEF, and Verna Lalbeharie Executive Director, EdTech Hub
Five Parts, Five Questions
Each part sets out what the evidence tells us, and what that means for practice.
1. Contextual Readiness
What evidence do we have about how digital and AI-powered tools can be implemented and scaled responsibly within real education systems?
2. Educational Impact
What evidence do we have about how digital technologies can have a meaningful impact on student learning and broader educational outcomes?
3. Inclusive Access
What evidence do we have about how digital and AI-enabled tools can support equitable access and participation for diverse learners?
4. Safety and Well-being
What evidence do we have about how digital technologies can be designed and governed to uphold children’s rights and protect their well-being?
5. Design for Learning and Teaching
What evidence do we have about how digital and AI-enabled technologies can support effective teaching and learning for greater equity?
Five Recommendations for the Sector
As a top priority for the whole education sector, there needs to be a concerted effort to establish the norms and standards that guide effective practice for digital and AI technologies in education.
Achieving lasting impact means reckoning with the commercial agendas that shape much of the EdTech market. Loss-leading products and subsidised access can embed dependence on particular vendors long before their true costs, licensing terms and lock-in risks become clear. This is not a case against AI in education, but for it: protecting future learners, teachers and systems means designing for financial sustainability and asking hard questions about vendor lock-in from the outset.
Drawing on contributions across the report, these five cross-cutting recommendations set out a course of action towards safe, inclusive, evidence-backed uses of digital and AI-powered technologies in education systems.
Equity
Make learning equity the defining driver of investment in digital and AI in education. Interventions should be designed for equity from the outset, co-created with communities, and assessed on whether outcomes are improving for the groups furthest behind, with investment in mother-tongue multilingual learning and continuous teacher development.
Safety
Enforce safeguarding and privacy requirements before digital and AI-powered tools reach schools, and keep monitoring safety and well-being throughout use. This means safety-by-design, child rights impact assessments, high-privacy defaults, minimised data collection and no targeted advertising, with contracted companies required to evidence their impact on children over time.
Pedagogy
Engage teachers as decision-makers and co-designers of digital and AI-enabled tools, so that technology serves pedagogy rather than displacing it. This means prioritising inclusive, culturally appropriate solutions and sustained investment in teacher professional development, financed through core education budgets and aligned with approaches such as structured pedagogy and Teaching at the Right Level.
Sustainability
Ensure contextual fit and cost-effectiveness are prerequisites for scaling, reflected in procurement, distribution and implementation choices that centre equity and protect the public interest. Design should focus on what works, for whom and under what conditions, without assuming one device per child or reliable connectivity.
Evidence
Invest in the human and institutional capacity to link evidence, innovation and implementation in a continuous cycle, built into government delivery rather than treated as a one-off exercise. This means focusing on educational outcomes from the outset, aligning with government planning and budgets, and staying attentive to political economy and enabling environments.
Global education is at a critical moment. Major changes are taking place, quickly, often without being well thought through and with significant vested interests. For all of us involved in how digital technologies are used in education it is time to lift our ambition, navigate with courage and make evidence-based decisions.”
– David Hollow, Research Director, EdTech Hub and Frank van Cappelle, Global Lead for Digital Education and Head of the UNICEF Global Learning Innovation Hub, UNICEF
Built in Partnership with
This report is the outcome of a collaboration between UNICEF’s Global Practice for Education, EdTech Hub and UNICEF’s Office of Strategy and Evidence. It draws on evidence reviews, think pieces and country case studies from a wide community of researchers, evaluators, practitioners and UNICEF country offices across the digital education ecosystem. It was produced under the editorial oversight of Juliette Norrmén-Smith and Iona Wotton, and the management of Frank van Cappelle, David Hollow and Thomas Dreesen.
EdTech Hub’s work is made possible with support from UK International Development.
