Source/event date: 23 October 2026
Published and reviewed: 26 October 2026
Scope: England
The Education Endowment Foundation has published new detail on how AI and other education technologies will be evaluated through England’s EdTech Testbed Programme.
The first cycle will test about 15 tools between October 2026 and April 2027. Promising approaches may progress to larger evaluations from spring 2027. The proposed evaluation guide will cover trial design, outcome selection, implementation and work with developers.
Importantly, the tender distinguishes platform use from educational impact. It says usage data can help explain adoption and engagement, but should not be treated as an ultimate outcome on its own. Evaluations may examine immediate performance and near transfer, alongside longer-term outcomes where feasible. Product changes during a trial must also be documented.
This is an evaluation framework under development, not a new requirement for schools and not evidence that any particular AI product works. For educators, it offers a useful set of questions: what outcome is being claimed, how will it be measured, what comparison is fair, and does any benefit persist beyond use of the tool?
Primary sources: EEF tender page and full invitation to tender (PDF).