Where Education Philanthropy Is Investing: Lessons from Overdeck's Q3 2026 Grants

1. AI — but tested, not assumed

The strongest theme is AI paired with evaluation. Rather than funding AI tools alone, many grants pay for research into whether they work: a national Brookings initiative on how uses of AI affect learning and trust, a University of Chicago study on when technology can substitute for human intervention, a RAND randomised trial of an AI attendance platform, and studies of AI writing feedback and maths simulations ^[1].

New pilot grants to companies such as Coursemojo, Timely and OKO Labs are framed as tests with lessons shared publicly ^[1].

2. Evidence-building as a funding category in itself

Roughly half the itemised grants went to research and evaluation organisations — WestEd, MDRC, RAND, AIR, Mathematica, Johns Hopkins and Stanford among them ^[1]. Overdeck groups these as 'ecosystem' grants intended to clear the path to scale for its direct-impact grantees ^[1]. This reflects the foundation's own strategy; it isn't evidence that funders generally favour research over services.

3. Mathematics and instructional materials

Maths runs through the largest grants: $3 million to Zearn for Louisiana Lift, with a separate WestEd evaluation of the same initiative, plus support for maths teacher simulations and curriculum-linked tools ^[1]. Several grants focus on implementing high-quality instructional materials well, rather than adopting new ones.

4. Attendance and family engagement

Two randomised trials target attendance — TalkingPoints (AIR) and Edia (RAND) — and a University of Chicago lab is testing an AI parent-engagement tool for school readiness ^[1].

5. Early childhood and school readiness

Early learning is a consistent priority: LENA's early-language programme, Collaborative Classroom's Pre-K portfolio, Pre-K assessment and curriculum studies by MDRC and the University of Michigan, and a national Kindergarten Readiness Assessment database ^[1].

What this means for schools

Overdeck does not accept unsolicited proposals ^[2], so schools cannot apply for these grants directly. The practical value is as a signal: if a project touches AI, maths, attendance or early learning, framing it around evidence and measurable outcomes is likely to resonate with similar evidence-focused funders. Partnering with organisations already funded in these areas may also open doors.

These conclusions describe one foundation's stated priorities for one quarter. Broader conclusions about education philanthropy need data from many funders over time, which K-12 Funding's Funding Intelligence dataset is starting to collect.