Digital Promise K-12 AI Infrastructure Program — Integrating Digital Public Goods to Advance AI-Enabled Formative Assessment
Open competitive grant: Proposals are due November 20, 2026, at 11:59 p.m. Anywhere on Earth (AoE). Digital Promise and its K-12 AI Infrastructure Program partners invite U.S.-based teams to produce openly licensed infrastructure for using education-specific data, benchmarks and models in AI-enabled K-12 formative assessment.
Individual award: Up to $600,000 per applicant. Up to two awards will be supported from a separate $1,000,000 total funding pool. Estimated performance: 18 months in three six-month phases, beginning February 1, 2027. The pool is not the individual award ceiling.
Who can apply
Eligible lead applicants are U.S.-based entities, including applicants in U.S. territories, edtech companies, other for-profit companies, nonprofit organizations (including fiscally sponsored nonprofits), school districts, and colleges or universities. The project outputs must clearly benefit U.S.-based students. One proposal per standard lead organization; organizations with over 2,000 staff may submit one per distinct department, and universities with over 2,000 staff may submit one per distinct major academic division or school. Teams should combine educational learning sciences, AI/software and community-engagement expertise.
What the grant supports
Funded teams will build an open-source package of developer tools, documentation, workflows and evaluations to help teachers and AI-enabled systems coordinate on formative assessment. In Phase 1, the proposal must cover at least four developer-facing tasks that meaningfully improve teacher/AI coordination. Later phases incorporate additional public goods, evaluate results, engage at least three downstream developers and release an openly licensed package. The majority of effort should serve multiple third-party developers, not fund a proprietary teacher-facing product. Code must use MIT or Apache 2.0 licensing; datasets and other assets must use CC-BY-4.0 as specified in the solicitation.
Allowable costs: Direct project staff/educator time, consultants, allocable equipment and computing, travel directly tied to project delivery, IRB and privacy oversight, essential software and tools, project subawards, translations and accessibility, and institutional overhead up to 15%. Disallowed: pre-award costs; lobbying or advocacy; entertainment; unrelated international travel; capital buildings/renovation; certain general administrative costs; and equipment above $5,000 without pre-approval and justification. No matching requirement or reimbursement/payment schedule was specified in the publicly reviewed RFP; applicants should confirm award terms before budgeting.
Application requirements
Submit the application through the official linked online Skipso form; registration is required. Include a 350-word abstract, a narrative PDF of no more than 6,000 words, a supplemental PDF with key personnel CVs and commitment letters when applicable, and a separate spreadsheet budget with justification using the prescribed template. Address significance, technical and learning-sciences expertise, workplan and milestones, developer adoption, safe and ethical AI, privacy, licensing and dissemination. Use the naming conventions and formatting in the controlling 19-page RFP and its FAQ. An application form template and budget spreadsheet are linked from the official program page.
Official controlling sources: RFP overview and application instructions, complete 19-page RFP PDF, and official FAQ. Contact the funder at k12ai-infrastructure@digitalpromise.org for clarifications.