One school now uses AI coaching to provide personalised feedback – without taking up mentors’ limited time. Professor Haili Hughes reveals whether others should follow suit
With PGCE placements now in full swing and early career teachers getting to grips with their own classes – mentors up and down the country are ready to support them.
Research consistently tells us mentors play an important role in helping schools support and develop new teachers.
Yet, a Teaching Commission report sounded the alarm on a challenge that goes beyond curriculum reform or accountability: England’s schools are facing a crisis in mentor capacity.
The commission found many schools are ‘stretched to breaking point’ and unable to release experienced staff from timetables to provide the sustained, one-to-one support that new teachers require.
In England, the early career framework has rightly raised expectations for the quality of mentoring.
But as an evaluation by the University of Gloucestershire recently found, the expansion of the early career framework has intensified workload pressures for mentors, with many juggling their teaching responsibilities alongside training and observation requirements.
The Education Policy Institute has also warned that without additional funding and staffing, “mentoring risks becoming tokenistic, with mentors too overstretched to provide meaningful feedback”.
Similar findings from the Sutton Trust underline that structured mentoring’s success depends on capacity, and capacity is exactly what schools lack.
Could AI coaching fill the gap?
If human mentors are in short supply, could artificial intelligence (AI) help plug the gap?
A promising avenue for using AI for one part of a mentor’s role – improving teaching – is AI-powered coaching tools.
At Church Hill Primary, part of the Central Region Schools Trust, colleagues have experimented with an AI coaching model as a way of providing regular, personalised feedback – without relying on senior leaders’ limited time.
Teachers essentially record their lessons, feed the transcript into an AI coach and receive immediate, private insights that focus on strengths and areas for growth.
This model has two advantages. It can provide every teacher with access to feedback every week, something impossible with human-only systems given staffing constraints.
And teachers report feeling less judged and more willing to engage with feedback when it comes from a neutral tool, rather than a line manager.
At Church Hill, nearly all staff have engaged enthusiastically with AI coaching – voluntarily recording multiple lessons per week and reporting increased confidence in their teaching.
Research from the Chartered College of Teaching notes that technologies which ‘democratise access to coaching’ could play a role in teacher retention, especially if they help new teachers feel less isolated.
AI mentoring and coaching could address equity concerns, providing consistency across schools where human mentoring capacity is uneven.
The drawbacks: What AI can’t do
But while the promise of AI is real, so too are its limitations. Mentoring is not just about pointing out patterns or areas for improvement. It is also about human connection and the important relational trust built over time.
The Teaching Commission report stresses that mentoring is as much about emotional support as it is about instructional guidance.
This is something that AI, at least for now, cannot replicate.
Towards a balanced model?
While AI may not be the full solution to the mentor capacity crisis, it can certainly be part of it. A balanced model could involve AI handling the scalable, low-stakes feedback loop, enabling teachers to reflect regularly and track progress.
Human mentors, freed from some of the burden, could then focus on the relational and developmental aspects of support that technology cannot provide.
Dismissing the potential of AI would be a mistake.
Without bold solutions, the ECF risks collapsing under its own weight – leaving early career teachers unsupported and schools unable to stem attrition.
The challenge for policymakers and school leaders is to resist both extremes: neither imagining AI as a silver bullet, nor dismissing it as irrelevant.
Opportunity lies in seeing AI as one tool among many, helping schools to see the wood for the trees and ensuring every teacher receives the support they deserve.
Your thoughts