STAR21 Submits Evidence to Parliamentary Inquiry on AI in Education

The Education Select Committee has launched a new inquiry examining how artificial intelligence and EdTech are reshaping education across England. The Committee is seeking written evidence on the benefits, challenges and risks of AI adoption in schools, colleges and universities, with a deadline of 10 April 2026.

STAR21 co-founder Adam Sturdee, who is also a serving assistant headteacher, has submitted written evidence drawing on his dual perspective as a school leader and as a practitioner who has designed and deployed an AI coaching tool in schools.

What the inquiry is asking

The Committee's terms of reference cover significant ground: the opportunities and challenges of AI in education, the impact on teaching practice and workload, data protection and children's digital rights, and whether the Government has an adequate regulatory framework. These are exactly the questions that have shaped the development of Starlight from the outset.

What we argued

Our submission addressed three core themes.

AI can solve the coaching frequency problem

The research evidence is clear that specific, timely, actionable and regular feedback improves teaching. The operational challenge has always been delivering that feedback at scale. Transcript-based AI coaching enables teachers to receive detailed, evidence-grounded reports after every lesson they choose to upload, without the scheduling, cover and write-up costs of formal observation.

Ethical design is not optional

The single greatest risk in this space is that AI tools designed to support teachers are repurposed for surveillance or performance management. Our submission set out six non-negotiable design principles: teacher-initiated recording, private-by-default feedback, no compulsory scoring, anonymised leadership data, teacher control over deletion, and explicit separation from appraisal. We believe these principles should be reflected in any Government guidance or quality standards for AI in education.

Schools need better support

From data protection impact assessments to evaluating the pedagogical claims of AI vendors, schools are navigating complex decisions without adequate guidance. We recommended that the DfE develop standardised DPIA templates for AI tools that process classroom audio, and that the Government establish quality standards to help schools distinguish between well-designed, evidence-informed tools and those making unsubstantiated claims.

Why this matters

The inquiry comes at a pivotal moment. The Government has invested £23 million in EdTech Testbeds and the Education Secretary has described AI as potentially transformative for education. At the same time, there is no coherent quality assurance framework for AI tools used in schools, and the line between coaching and compliance remains dangerously blurred in parts of the market.

Getting this right matters for teachers, for school leaders, and for the pupils who benefit when their teachers receive high-quality professional development. We are glad Parliament is asking these questions and we look forward to contributing to the conversation as the inquiry progresses.

Adam is presenting peer-reviewed research on transcript-based AI coaching at the BERA Teacher Education Advanced Network (TEAN) Conference at Sheffield Hallam University in May 2026.

If you are interested in how Starlight can support coaching at scale in your school or trust, you can book a demo here:

https://starlightmentor.com/demo-request