What a Stanford AI Coaching Trial Tells Us About the Future of Teacher Development

One of the most common questions we’re asked about Starlight is a fair one:

“Is there any serious research showing that AI feedback can actually improve teaching?”

The answer is yes — and some of the most compelling evidence comes from Stanford University.

In this post, we want to briefly outline a Stanford-led research trial and explain why its findings matter so directly for what we’re building with Starlight.

The Stanford Study in Brief

Researchers at Stanford ran a large-scale randomised controlled trial exploring whether automated, AI-generated feedback could improve how teachers respond to student ideas during lessons.

The study focused on a specific but powerful aspect of classroom practice:

uptake — the extent to which teachers build on, probe, or extend what students say, rather than simply evaluating it and moving on.

Using transcript analysis and natural language processing, the researchers:

• Analysed teaching transcripts at scale

• Generated automated feedback for instructors

• Compared outcomes between an AI-feedback group and a control group

Importantly, this feedback was:

• Non-judgemental

• Specific

• Grounded in real lesson data

• Delivered without a human coach

What Did They Find?

The results were striking.

Teachers who received AI-generated feedback showed:

• A significant increase in uptake of student ideas

• A shift away from dominant teacher talk

• Measurable improvements in learner experience and engagement

All of this happened without observations, grading, or performance management — just structured insight drawn from transcripts.

This matters because it demonstrates something crucial:

Transcript-based AI feedback can change teaching behaviour.

Not in theory. Not as a pilot idea.

But in a controlled trial, at scale.

Why This Matters for Starlight

Although the Stanford study took place in an online teaching context, the underlying mechanism is exactly what Starlight is designed to do in schools.

At the heart of both approaches is the same idea:

• Real lesson data

• Analysed carefully and ethically

• Returned to teachers as private, reflective insight

This aligns closely with established research on dialogic teaching, classroom talk, and the long-standing limitations of traditional IRF (Initiation–Response–Feedback) patterns when overused.

What Stanford has shown is that AI can help teachers notice their own patterns of talk, and that noticing — when done carefully — leads to meaningful professional growth.

That insight sits right at the centre of Starlight.

Read the Paper Yourself

We strongly encourage anyone interested to read the original paper.

You can access the full Stanford publication here:

👉 https://scale.stanford.edu/sites/default/files/ai24_1054_v2.pdf

It’s rigorous, measured, and refreshingly cautious — which is exactly how educational research should be.

A Final Thought

Starlight isn’t about replacing coaching, judgement, or professional dialogue.

It’s about making insight visible, at a scale and frequency that schools have never been able to sustain before.

Stanford’s research gives us confidence that this approach is not only plausible — it’s already working.

And now, we’re building it for schools.

Spark Insight with Starlight — and inform practice today.

🎥 Subscribe to our channel here: https://www.youtube.com/@Star21-ai

🌐 Read more on our blog: www.coaching.software

💡 Explore the platform: www.starlightmentor.com

🐦 Follow us on X: @star21starlight

The Insight Engine is written by Adam Sturdee, co-founder of Starlight—the UK’s first AI-powered coaching platform—and Assistant Headteacher at St Augustine’s Catholic College. This blog is part of a wider mission to support educators through meaningful reflection, not performance metrics. It documents the journey of building Starlight from the ground up, and explores how AI, when shaped with care, can reduce workload, surface insight, and help teachers think more deeply about their practice. Rooted in the belief that growth should be private, professional, and purposeful, The Insight Engine offers ideas and stories that put insight—not judgment—at the centre of development.