The Estonian AI leap: navigating uncertainty through the production of education futures

This article analyses the Estonian AI Leap initiative as a contemporary production site for education futures. Drawing on the concept of ‘refraction’, the study examines how a shared vision of artificial intelligence as an ‘arrival technology’ with uncertain effects on education is negotiated and reshaped by a sociotechnical vanguard composed of policy actors, researchers, and technology entrepreneurs. Based on expert interviews with these actors and media coverage of the project, distinct refractions through which the uncertainty surrounding AI in education is managed are identified. The first emphasizes speed and global competitiveness and is oriented towards short iterations and actionable knowledge. The second reframes AI through evidence-based approaches to learning, positioning scientific authority as a means of regaining control and mitigating harm in the high-stakes context of education. The third articulates a practice-based approach, which tries to balance the possibilities of new technologies with the need to reassert control in the face of possible harmful effects. The ways in which uncertainties are handled across the three refractions are underpinned by distinct epistemic imaginaries. Each epistemic imaginary privileges particular forms of knowledge and expertise, which provide legitimization to the specific claims concerning how to steer educational trajectories toward desirable education futures.