Based on a survey of 1,660 teachers teaching 5-18 year olds across ten countries, Futuresource’s latest whitepaper, The Impact of AI on Teaching and Learning: K-12 Outcomes, Evidence and Enabling Factors, finds that pedagogy, teacher training, infrastructure and governance have a greater influence on educational impact than AI technology itself.
For policymakers and EdTech suppliers, the message is clear: successful AI implementation requires more than deploying tools. Investment in teacher training, pedagogy, infrastructure and clear governance frameworks will be critical to achieving meaningful impact at scale.
While 50 per cent of teachers report positive effects on student learning, only 13 per cent describe AI’s impact as “very positive”, highlighting a significant gap between adoption and meaningful outcomes.
One of the study’s strongest findings concerns teacher capability. Approximately 75 per cent of teachers who feel well trained in AI report positive impacts on teaching and learning, compared with just 38 per cent of those with little or no training. Yet more than a quarter of teachers report receiving no AI-related training at all.
“Much of the discussion around AI in education focuses on the technology,” said Claire Kerrison, principal analyst at Futuresource Consulting and author of the report. “Our research shows that AI acts as an amplifier of existing teaching practice. It can strengthen great teaching, but it cannot compensate for weak implementation.”
The report challenges another common assumption: that increased screen time is inherently harmful to learning.
Teachers working in 1:1 device environments were more than twice as likely to report significant improvements in learning outcomes than those in classrooms where devices were shared among multiple students. 36 per cent of teachers in 1:1 settings reported significant improvements, compared with just 15 per cent where devices were shared between four or more students.
“The findings suggest that the issue is not screen time itself,” said Kerrison. “What matters is whether students and teachers have reliable access to technology that can be integrated meaningfully into teaching and learning.”
The study also reveals that teachers in developing education systems often report stronger positive impacts from AI than those in more mature markets. For example, 17 per cent of teachers in Brazil reported AI having a very positive impact on day-to-day teaching, compared with just 3 per cent in Singapore. The findings suggest AI could help reduce disparities in educational quality by expanding access to instructional support and learning resources.
As education systems move from experimentation to implementation, Futuresource believes the conversation must shift from technological capability to educational outcomes.
