Lessons from the School Teams AI Collaborative

When I consider AI’s ideal role in education, I circle back to a simple truth: great educators drive great learning. AI can accelerate or amplify that work, but it will never replace the relationships, judgment, and humanity at the core of teaching.

That’s why we launched the School Teams AI Collaborative with our partners at FullScale. We wanted to move past the hype and fear around AI and ask a more practical question: What happens when whole school teams–teachers and leaders together–experiment with AI in service of student learning?

Over the past year, nearly 80 educators across 19 schools dug into that question. Their stories are as diverse as their contexts, but a common theme emerged: when school teams had the space, support, and trust to explore, AI became a tool for meaningful instructional improvement.

Beyond Hype: What We Learned

The public conversation about AI in schools has mostly centered on extremes: either breathless promises of revolution or dire warnings about cheating and risk. The Collaborative revealed something more grounded and more hopeful.

At the Eliot School in Boston, teachers used AI to generate student feedback, then built reflection routines so that feedback actually deepened learning. At Lynwood High School in California, a custom chatbot supported multilingual learners in accessing complex texts. At DSST Public Schools in Denver, students used AI to design chatbots that solved real problems in their communities.

These weren’t flashy experiments. They were practical, thoughtful moves to do three things that matter most in classrooms:

• Do More: free up time for deeper feedback and planning.
• Do Better: sharpen instructional practices like project-based learning.
• Do New: design opportunities that hadn’t been possible before, like student-led civic problem-solving.

Taken together, these examples show that AI can serve learning but only when anchored in a clear instructional vision and guided by educators working in teams. Want to carry the learning forward? We and FullScale created ready-to-implement strategies that reflect what early adopters are trying.

The Conditions That Matter

One of the clearest lessons from this work is that AI won’t transform anything on its own. To unlock its potential, schools and systems need the right conditions in place:

1. A shared instructional vision so AI use is coherent, not scattered.
2. Cross-functional leadership structures that connect academics, technology, and professional learning.
3. Investment in teacher capacity through sustained, job-embedded professional learning.
4. Policies that treat AI as an instructional lever. Tech policies often stop at data privacy, procurement, or safe use; with AI now woven into pedagogy, systems must set policies that align to the instructional vision and goals, ensuring it’s clear how AI strengthens teaching and learning, not just how it gets purchased or deployed.

These are not “nice-to-haves.” They’re the difference between schools reinforcing the status quo or taking the step to ask, “Are we ready to pursue meaningful change, or are we more comfortable upholding what we already know?”

Why Practitioner Voice Must Lead

Too often, the people shaping policy and building products for classrooms and schools do so without listening to the people actually in them. The Collaborative showed that educators are not passive implementers; they are critical co-creators of how AI can serve students. 

That matters because too many edtech products solve problems teachers don’t actually have (or solve problems that aren’t critical and high leverage for teachers and students).

• Despite the abundance of tools, teachers aren’t reporting a decrease in the challenges they face.
• By contrast, Collaborative participants surfaced ideas grounded in daily practice and aligned to their instructional priorities: feedback loops that worked with existing routines, supports for multilingual learners, and building real-world skills like collaboration, independence, and more.

If we want AI to strengthen learning, educators’ insights must shape the tools we build, the policies we adopt, and the investments we make.

What Comes Next

Today, we are releasing Generative Practice: Practical Insights for Unlocking the Instructional Potential of AI, a report that shares direct learnings from and with practitioners, along with steps system leaders can take to create the conditions for impact.  This report is not an endpoint; it’s a foundation.

At Leading Educators, we believe coherence is critical. It helps bridge innovation and daily practice, making sure that new tools and ideas like AI don’t live on the margins but are integrated into curriculum, professional learning, and system priorities. Without coherence, innovations stay isolated. With it, they strengthen the core of teaching and learning.

Here’s how those lessons are already shaping our next steps:

• Leveraging AI in curriculum implementation: We are piloting curriculum-based professional learning and coaching approaches that integrate HQIM-specific AI tools and routines for math, and we are in the early stages of developing similar supports for ELA.
• Developing solutions for various contexts and spheres of influence: We recently launched the RAISE AI Collaborative in partnership with aiEDU, Collegiate Edu-Nation, and the Arizona Institute for Education and the Economy. This collaboration builds on the adaptability and creativity that rural schools bring to testing new approaches to learning.
• Future-ready leading and learning: This summer, we tested new approaches to teaching and learning that foreground durable skills, student agency, and leveraging AI; we are gathering early signals that will inform how professional learning systems, what educators need, and how student learning experiences must evolve in an AI-enabled world so that LE and others can prepare not just for what’s now but what’s next.

Each of these efforts builds on the same insight: innovation sticks when it starts with educators, is anchored in coherence, and serves student learning.

Gratitude and An Invitation

We are deeply grateful to the school teams that opened their classrooms and shared their learning with us this year; their collective leadership and the work school teams do each day are paving the way for more thoughtful, intentional, and impactful integration of AI in teaching and learning. We also thank our partners at FullScale, who helped us figure out how to capture, make meaning of, and share our collective learning from the year, creating different ways to ensure our collaborative efforts don’t end here.

I hope that this report doesn’t sit on a shelf. I hope it sparks honest conversations in schools, districts, and states about how we prepare for an AI-enabled future. While the School Teams AI Collaborative was designed to surface early practices and insights, we know that’s only the beginning. There is more to explore, test, and build if AI is going to serve as a lever for deeper learning.

I also hope it challenges us, as leaders, funders, and partners, to put educators at the center. If AI will reshape teaching and learning, let’s make sure it does so in ways that expand opportunity, build student agency, and strengthen the core of what great teachers already do every day.

AI won’t save schools. But in the hands of school teams, it can help us take meaningful steps toward something better.