2025 AI Strategy Update

This post is authored by Anu Malipatil, Executive Director, and Josh Lotstein, Senior Director, Grantmaking.

The latest NAEP results send a clear and urgent message: students need more support to recover and accelerate their learning, and our education system needs more effective solutions to help them do so. Since 2019, reading scores have dropped across the board for both fourth and eighth graders, and a staggering one-third of eighth graders are now reading below the basic level. These numbers aren’t just statistics—they represent students who are struggling, classrooms where learning isn’t as effective as it could be, and teachers who need better tools to help their students succeed.

At Overdeck Family Foundation, we believe artificial intelligence (AI) has the potential to be one of those tools—not as a replacement for teachers, but as a powerful assistant to help educators make learning more engaging, personalized, and effective. Our updated AI strategy is grounded in a simple but critical goal: to understand what works to improve teaching and learning in America’s schools.

We’re not interested in AI for AI’s sake. Instead, we’re focused on measuring efficacy (does it improve learning?), efficiency (does it make effective teaching more scalable?), and cost-effectiveness (is it a good use of public funding?). We want to ensure that AI systems, programs, and applications aren’t just another distraction in our schools, but offer real solutions that make a meaningful impact on student outcomes.

What we’ve learned so far

After a year of investing in and observing AI applications in education, we’ve learned a few key lessons, which we detailed in a previous blog post. Here are some of the most meaningful takeaways that surfaced:

• AI is mostly being used to help educators, not yet students: Right now, most AI adoption in education is happening behind the scenes—helping teachers plan lessons, manage administrative tasks, and streamline workflows. While this support is valuable, we see an opportunity to responsibly push further into AI-powered tools that directly improve student engagement and learning.
• AI coaching for teachers shows real promise: AI-driven coaching tools are starting to gain traction in professional development. Early research suggests that these tools can help teachers improve their instructional practices by offering real-time, objective feedback that’s cost-effective compared to traditional coaching methods.
• Personalized learning through AI needs more evidence: AI’s potential to customize content, assessments, and support for individual students is one of the most exciting possibilities in education. But while the idea is promising, we still need more research to understand what really works at scale. The key question isn’t just whether AI can personalize learning, but whether that personalization actually improves outcomes for the medium and long term.

These insights have informed our Foundation’s 2025 AI investment strategy and its focus on understanding the efficacy, efficiency, and cost-effectiveness of AI for teachers and students. We are also grateful for the added input and perspective from experts in the field, ranging from researchers to investors and practitioners. A huge thank you to Amber Oliver, Kumar Garg, Ben Kornell, Robin Lake, Shalinee Sharma, Ben Wellington, and Chris Mulligan.

What you’ll see in 2025

As we refine our investment approach in AI, we’re focusing on four major areas:

1. Funding AI tools that address real education needs and use cases: Prioritizing AI solutions for assessment, content personalization, and administrative efficiency to improve both student learning and teacher sustainability.
2. Conducting research on AI’s impact: Funding studies to measure how AI tools can improve student outcomes, educator efficiency, and cost-effective spending.
3. Building an AI education ecosystem: Developing benchmarks to evaluate AI models in education and underlying data sets, as well as the tools that use them. In addition, investing in approaches and solutions that create strategic alliances between nonprofit and for-profit partners.
4. Helping nonprofits adopt AI thoughtfully: Supporting grantees in using AI effectively, not just for technical skills but for collaborative learning and innovation.
   Guiding all these investments is the underlying belief that AI has the potential to improve teaching and learning, reduce inefficiencies, and create new opportunities for students and educators alike—but only if we’re strategic about what and how it’s implemented.

A call to action

The time to act is now. The NAEP results make clear that we need to rethink how we support students and teachers in order to reach the outcomes we all believe are possible. AI is one tool that could help. But realizing its potential will take more than just developing the technology—it requires investment, collaboration, and rigorous testing.

We invite funders, researchers, developers, and educators to join us in this work. Whether through co-funding high-impact AI solutions, partnering on research, or sharing best practices, we believe that a coordinated effort will help us better understand and use AI to its full potential.

If you share our vision of AI as a force for better learning experiences, smarter teaching tools, and improved education outcomes, we invite you to reach out to our team.

Header image courtesy of Saga Education