Does guided use of Microsoft Copilot reduce the time teachers spend on lesson planning in 16 – 19 education?

What is the project?

This project explores whether receiving practical guidance on using Microsoft Copilot can help teachers in further education (FE) or sixth form colleges reduce the time they spend planning lessons without reducing the quality of teaching materials.

The project centres on a short digital Copilot guidance resource developed by the Behavioural Insights Team (BIT), with advice from academics and FE practitioners. Participating subject departments will be randomly allocated to either receive the Copilot guidance or continue with their usual lesson-planning approaches during the trial (including the use of other GenAI tools). Researchers will compare the time spent on lesson planning and the quality of lesson resources between the two groups using weekly diaries, teacher surveys, lesson observations, lesson materials and qualitative feedback.

Who is leading this project?

The trial is funded by the EEF and delivered and evaluated by BIT. BIT developed the Copilot guidance, manages the evaluation, and provides support to participating colleges throughout the project. The Principal Investigator is Dr Patrick Taylor.

What will this project look like in your setting?

Participating colleges will nominate at least five eligible subject departments from the following list (or similar): English GCSE resits; maths GCSE resits; health and care services; business management and administration; engineering and manufacturing; education, training and childcare; crafts, creative arts and design; media, broadcast and production; public services; sport, leisure and recreation; construction, planning and the built environment. They will also identify a Research Champion to coordinate activity within each college site.

Departments will be randomly assigned to one of two groups:

1. Intervention groupAs part of a Randomised Controlled Trial (RCT), settings will be randomised into either the intervention or control group. Settings in the intervention group will receive the programme being tested. (Copilot guidance): Teachers receive access to BIT’s digital Microsoft Copilot guidance from w/​c 1 February 2027.
2. Control groupAs part of a Randomised Controlled Trial (RCT), settings will be randomised into either the intervention or control group. Settings in the control group continue with their usual practices and help provide a comparison to measure the intervention’s impact. They are usually offered a monetary compensation as thanks for their contribution. (carry on as usual): Teachers continue their normal lesson-planning practices and receive access to the Copilot guidance after the trial ends on 1 June 2027.

All participating teachers will complete an online survey in January before the trial starts. They will then complete weekly diaries from 1 February to 21 May 2027, followed by an end-of-trial survey at the end of May 2027. The weekly diaries will take fewer than 5 minutes to complete and will capture AI use and teacher time spent on lesson planning. Around half of teachers will also be asked to share anonymised lesson materials to assess quality. A small number of colleges may be invited to participate in interviews, observations or learner focus groups.

There is no cost to participate. Teachers who complete the first survey in January can receive up to £75 in vouchers for completing diary and survey activities, with additional incentives available for sharing lesson materials or taking part in interviews. Participating college sites will also receive a £100 providing the teachers who receive the Copilot guide do not share it with others during the trial.

Who can take part?

To participate, colleges must:

• Be a general FE college or sixth form college in England;
• Nominate at least five eligible subject departments teaching 16 – 19 learners (from this list (or similar): English GCSE resits; maths GCSE resits; health and care services; business management and administration; engineering and manufacturing; education, training and childcare; crafts, creative arts and design; media, broadcast and production; public services; sport, leisure and recreation; construction, planning and the built environment);
• Ensure participating staff can access and use Microsoft Copilot Chat through their college Microsoft 365 account;
• Have an AI policy that permits the use of Copilot for lesson planning;
• Appoint a Research Champion;
• Not nominate departments that are participating in another EEF trial involving the same learners or subject areas during 2026/27.

How can you register your interest?

Complete the Expression of Interest form here. The project team will then contact you with further information about eligibility, next steps and participation requirements. Colleges selected for the trial will be asked to submit a signed Memorandum of Understanding (MOU) and a technical readiness form.

Artificial intelligence (AI) has the potential to reduce teacher workload, a significant challenge across the 16 – 19 sector. Previous EEF Teacher Choices evaluations in schools have found that AI tools can substantially reduce lesson-planning time while maintaining the quality of teaching resources. However, evidence also suggests that sustained adoption can be difficult, with many teachers lacking the confidence, expertise or practical guidance needed to use AI tools consistently and effectively. In FE and sixth form colleges, Microsoft Copilot is already widely used, but there is very limited causal evidence on its impact on teacher workload, quality of outputs, or the conditions needed for successful implementation.

This project will generate robust evidence on whether guided use of AI can meaningfully reduce time spent on lesson planning without compromising quality. It will also test whether structured implementation guidance can improve adoption and effective use of AI in lesson planning. The findings of this study will help colleges, policymakers and sector leaders make informed decisions about the role of AI in supporting teachers and lecturers in the 16 – 19 sector.

The project will be evaluated by the BIT using a two-armed randomised controlled trialAn RCT is used evaluate an educational programme by assigning settings to one of two groups: the intervention group, who receive the programme or the control group, who continue with business as usual. This ensures that any differences in outcomes can be confidently attributed to the programme, providing a robust estimate of the impact and contributing to the evidence for what works in improving educational outcomes. (RCTAn RCT is used evaluate an educational programme by assigning settings to one of two groups: the intervention group, who receive the programme or the control group, who continue with business as usual. This ensures that any differences in outcomes can be confidently attributed to the programme, providing a robust estimate of the impact and contributing to the evidence for what works in improving educational outcomes.). This is a Teacher Choices trial that tests the daily practical approaches teachers use to support teaching, using rigorous research methods to understand their impact on teachers’ practices and outcomes for teacher workload.

Participating subject departments will be randomly allocated to either receive guidance on how to use Microsoft Copilot for lesson planning or continue with business-as-usual practice. The evaluation will compare the two groups on teacher time spent on lesson planning and the quality of lesson materials. AI use and teacher confidence will also be collected. An implementation and process evaluationAn IPE is used to understand how and why an intervention has (or has not) been successful. Data is analysed to explore programme quality, reach, adaptation and differentiation, as well as setting fidelity and responsiveness to the trial design. will explore how the Copilot guidance is adopted and used in practice.