Sorry, marketing cookies are required to view this form.
Jul 29, 2025 · 2 min read
Building Trust in AI for Education
CZI is providing edtech companies with resources to help make their AI-powered tools more rigorous, more accurate, and better aligned with what and how students learn.
We have all used tools with gen AI, and while they may sometimes sound smart, we know they’re not always totally right,” says Sandra Liu Huang, CZI’s head of education and vice president of product. “So how do we raise our confidence in this technology to a level that’s worthy of our teachers and of our students?”
It’s a question at the heart of CZI’s work to improve the quality of edtech tools.
For nearly a decade, we’ve worked to advance learning science research and help translate that research into classroom practice. With the advent of generative AI, that translation work can be accelerated and scaled to have even greater impact, but there are fundamental technical and infrastructural resources that need to be built in order to help AI-powered edtech tools be truly impactful.
In this video, Huang joins Yusuf Ahmad, CEO of Playlab, and Michelle Odemwingie, CEO of Achievement Network, to share how CZI is providing edtech companies with tools and resources to help make their AI-powered products more rigorous, accurate, and better aligned with what and how students learn.
Also read: Knowledge Graph Helps Build AI Rooted in Learning Science
The first of these resources is Knowledge Graph, which has interconnected data sets that bring together high-quality curricula, state academic standards, skills and learning science research to help edtech developers ground their products in the context of the classroom. The second is Evaluators, which enables developers to ensure high-quality AI-generated content by assessing outputs against expert-backed rubrics.
Before CZI’s Knowledge Graph, we had educators manually uploading Illustrative Mathematics curricula and getting wildly unreliable results. Now, after months of iteration and testing Knowledge Graph within PlayLab, we’re seeing much higher quality outputs that are more accurate and more aligned to Illustrative Mathematics design principles.
Explore the video to learn how these tools are already making a difference and partner with us to advance AI in education for the public good.
