K-12 AI Infrastructure Program

K-12 AI Infrastructure Program is sponsored by Digital Promise (with funding from Bill & Melinda Gates Foundation). This multi-year initiative aims to close the gap between scientific principles of teaching and learning and the promise of generative AI.

Digital Promise Announces First Grantees of the K-12 AI Infrastructure Program – Digital Promise Digital Promise Announces First Grantees of the K-12 AI Infrastructure Program June 29, 2026 | By Digital Promise Digital Promise, alongside core partners Learning Data Insights , DrivenData , the Massive Data Institute at Georgetown University , and Catalyst @ Penn GSE , today announced the first grant recipients of the K-12 AI Infrastructure Program , a new $26 million multi-year initiative to improve how artificial intelligence (AI) is designed and used in K-12 classrooms.

The program focuses on a simple but critical challenge: while AI is rapidly entering schools, many tools are not yet designed in ways that reflect how students actually learn or how teachers assess understanding in real time. As a result, educators often struggle to trust or effectively use these tools.

“Without the right foundation, AI will become another barrier to educational progress,” said Jean-Claude Brizard, president and CEO of Digital Promise. “Through this work, we’re building the shared resources developers need to create AI tools that give teachers better insight into student learning and help students get the right support at the right time.

” This first grant cycle focuses on strengthening formative assessment – the everyday practices teachers use to understand what students know while learning is happening, such as classroom discussions, written work, or problem-solving activities. While formative assessment is one of the most effective ways to improve student outcomes, it can be difficult to implement consistently at scale.

The K-12 AI Infrastructure Program aims to make this easier by supporting the development of openly available datasets, benchmarks, and models that help AI tools better interpret student thinking and provide meaningful feedback. By making these resources public, the program lowers barriers for developers while also raising expectations for quality, accuracy, and fairness.

“A cornerstone of this work has been engagement with educators and community organizations to ensure our investments reflect their goals and concerns for AI in education,” said Rebecca Griffiths, program design director at Digital Promise.

“Educators want AI that understands context and empowers teachers to do their best work — and they’ve been clear that student privacy and agency must remain central to how these tools are designed and used. ” Following a competitive review process that brought together funders, researchers, developers, and educators, Digital Promise and its partners selected four organizations to receive grants for projects lasting six to 12 months.

All outputs will be openly licensed and freely available for use across the education field. The Digital Promise team will be making additional grants in coming months and years, with a total of approximately 30 grants expected in the life of the program.

Learning Equality, PI: Richard Tibbles A Benchmark for AI Identification of Science Misconceptions in Free-Form Student Responses Princeton University, PI: Tammy Kwan Pipeline for Training Simulated Student Models with Reduced Human Data National Tutoring Observatory/Cornell University, PI: Allison Koenecke Open Leaderboards for Benchmarking Automated Speech Recognition in Educational Contexts Stanford University, PI: Hariharan Subramonyam KB-TutorBench: A Multimodal Knowledge-Building Dataset for AI-Enabled Formative Assessment “The projects chosen for funding are applying advanced data science approaches, often proven in other fields, in education contexts,” said John Whitmer, senior researcher and founder of Learning Data Insights.

“It was a difficult selection given the depth and breadth of responses that we received, and I look forward to seeing the public goods emerge from these projects for the field. ” This initiative is supported by Learning Commons , Charles and Lynn Schusterman Family Philanthropies , Gates Foundation , Overdeck Family Foundation , Valhalla Foundation , and Walton Family Foundation .

These projects reflect the range of ways AI could better support teaching and learning. Each project is designed with a focus on students and communities who have been historically underserved. Over four years, the program will continue to fund the development of these shared public goods to improve quality, relevance, and trustworthiness of AI in education.

For more information, visit https://k12-ai-infrastructure. org/ .

Center for Learning Sciences Research Learning Sciences Research Center for Learning Sciences Research Learning Sciences Research Artificial Intelligence Research Trust, Language, and Learner Voice: 3 Signals From Badge Summit By Marilys Galindo and April Williamson Transforming Education: New Position Paper Introduces the Powerful Learning Framework New Grants Fuel Research into Math Education with Datasets from the University of Florida Lastinger Center for Learning Beyond the Screen Time Debate: Reclaiming EdTech ROI Through Intentional Procurement We want to hear from you!

Please take this 5-minute survey and help us serve you better. We want to hear from you! Please take this 5-minute survey and help us serve you better.

Activating the following search input element will open the search modal. League of Innovative Schools Verizon Innovative Learning Schools