NSF's FINDERS FOUNDRY Pulls A DARPA-Style Two-Phase Gate Into K-12 AI Education With $8.5M, 70 Awards, And A Mandatory Four-Stakeholder Team Structure. Development Proposals Are Due November 18.

The National Science Foundation published the solicitation for NSF 26-507, the Fostering Interdisciplinary Networks to Develop Emergent and Responsive Solutions Foundry — operationally referred to as FINDERS FOUNDRY — earlier this spring, with a Planning-proposal deadline of May 27, 2026 that has now passed and a Development-proposal deadline of November 18, 2026 that defines the rest of the program's first-year cohort. The program is a comparatively modest $8.5 million total commitment with 70 anticipated awards, but its design represents a structural shift in how NSF is approaching K-12 educational technology research, and the design choices it embeds — the two-phase Planning-to-Development gate, the four-stakeholder team composition mandate, the explicit translation pathway to Innovation Corps and SBIR/STTR — are likely to be reused in subsequent NSF education solicitations regardless of FINDERS FOUNDRY's eventual measured impact.

The program is also one of the most direct operational implementations of Executive Order 14277 on Advancing Artificial Intelligence Education for American Youth, signed earlier in the administration as the policy framework for federal investment in K-12 AI literacy and skills development. NSF has been the federal agency tasked with translating the EO's stated goals into actual grant-making mechanisms, and FINDERS FOUNDRY is the first solicitation that visibly incorporates the EO's emphasis on AI as both a subject of K-12 learning and a tool for K-12 instructional delivery. The program also draws explicit alignment to the CHIPS and Science Act of 2022, which established the broader federal commitment to expanding STEM and AI workforce development across K-12 and postsecondary pipelines.

For organizations watching NSF's K-12 education portfolio — university Schools of Education, district research-practice partnerships, AI-focused educational nonprofits, museums and informal science learning institutions, and the small but growing community of K-12 EdTech startups operating in the NSF-fundable research-translation zone — the operational question for the next six months is the November 18 Development deadline, and the strategic question is what FINDERS FOUNDRY's design signals about the next decade of NSF K-12 grant-making.

What the program funds and how the two phases work

FINDERS FOUNDRY operates as a two-phase staged investment structure with a hard gate between phases. The Planning phase awards up to $50,000 per project for up to two months of work, with an anticipated 50 Planning awards in the first cohort. The Development phase awards up to $300,000 per project for one year, with an anticipated 20 Development awards drawn exclusively from the pool of Planning awardees. The math implies a 40% advancement rate from Planning to Development, with the explicit understanding that NSF will use the Planning-phase deliverables as the basis for selecting the Development cohort.

The structure is borrowed, intentionally, from the DARPA program management model that has shaped advanced research agency funding for two decades: a small, fast, low-cost first stage that lets the agency surface and evaluate many concepts simultaneously, followed by a meaningfully larger investment in the subset of concepts that have demonstrated early viability. NSF has experimented with similar staged structures in its Convergence Accelerator and Translation to Practice programs, but FINDERS FOUNDRY is the first time the model has been applied to K-12 education research specifically, and the time compression on the Planning phase (two months) is notably tighter than the agency has used in prior staged programs.

Planning-phase deliverables, per the solicitation, are oriented toward concept validation: stakeholder engagement with the target K-12 population, data collection on the specific learning challenge the team is addressing, wireframe design for the proposed technology or instructional approach, early pilot testing of the wireframe with actual K-12 students, and a team-formation and collaboration plan that will govern the Development phase if the team advances. The Planning grants are not literature reviews. They are explicit requests for two-month sprints producing testable artifacts.

Development-phase deliverables are oriented toward prototype-to-MVP progression: building and refining a working prototype of the proposed innovation, iterative testing with K-12 learners, validation against measurable learning outcomes, and preparation for a public showcase event that NSF will convene in Washington at the end of the funding cycle. Development proposals must include Institutional Review Board (IRB) approval or exemption documentation for the K-12 student involvement that the program requires.

The structure of the funding and the orientation of the deliverables together communicate a strong NSF preference for applied, near-prototype work over basic research — a shift that will surprise some education researchers accustomed to NSF's Discovery Research K-12 (DRK-12) program and the Innovative Technology Experiences for Students and Teachers (ITEST) program, both of which have historically funded longer-horizon, less prototype-oriented research.

The four-stakeholder team mandate is the structural break

The most distinctive design choice in FINDERS FOUNDRY is the mandatory team composition requirement. The solicitation requires that every team include at least one member from each of four stakeholder groups:

1. K-12 educators — classroom teachers, instructional coaches, curriculum specialists, or district instructional leadership
2. Technologists — software engineers, AI/ML practitioners, learning engineers, or EdTech product designers
3. Researchers — learning scientists, educational researchers, or domain researchers in STEM/AI
4. Parents or guardians of K-12 students

The Principal Investigator must be affiliated with an eligible organization (universities, nonprofits, museums, state/local government, federally recognized Tribal Nations), but the team itself must be cross-disciplinary by construction. This is a meaningful departure from how NSF has historically structured education research grants, where teams typically consisted of researchers from one or more universities with practitioner advisory boards or research-practice partnerships in supporting roles. The four-stakeholder mandate elevates parents/guardians to a co-equal team membership role — not advisors, not focus group participants, team members.

The rationale, per the solicitation language, is to ensure that the innovations developed under FINDERS FOUNDRY address challenges identified by K-12 students, families, and educators themselves, rather than challenges identified by researchers and technologists working at a distance from classrooms. The agency is explicitly trying to break the well-documented pattern in education R&D where technologies developed in university labs fail to translate into classroom use because the development process did not adequately involve the people who would have to use them.

The operational implications of the mandate are significant. University-based PIs cannot simply assemble a team of co-PIs from other universities and add a teacher advisory board. They must recruit, formalize, and compensate at least one classroom educator, one industry technologist, and one parent/guardian as co-equal team members with clear roles in the project work plan. Many universities do not have institutional infrastructure for this kind of cross-sector team contracting — particularly the parent/guardian role, which most university research administration offices have never structured a research subcontract around.

The mandate also requires that K-12 students themselves be involved throughout the development process — "in idea generation, design, testing and improving the technology," per the solicitation. This is not a passive research-subject relationship; it is a co-design relationship that requires IRB-approved participation structures, parent/guardian consent, age-appropriate compensation or recognition, and meaningful integration of student feedback into design decisions. Few research universities are operationally equipped to run K-12 co-design at this depth, and the teams that have done so successfully — typically through long-running research-practice partnerships with school districts — will have a significant structural advantage in the Development competition.

Eligibility and submission constraints

The eligible-organization list is broader than most NSF research solicitations: U.S. institutions of higher education (two-year and four-year, including community colleges), non-profit non-academic organizations (museums, research labs, professional societies), state and local governments, and federally recognized Tribal Nations. The inclusion of community colleges is notable — NSF's research portfolio has historically been dominated by R1 universities, and the explicit eligibility of two-year institutions signals a deliberate effort to expand the institutional base of NSF education R&D.

Each organization may submit one Planning proposal per PI/co-PI, and organizations that received Planning awards may submit one Development proposal per PI/co-PI. The submission limits are restrictive enough that large universities will need to internally coordinate which faculty members are submitting under the institution's quota.

The funding structure has additional operational constraints worth noting:

• No cost-sharing is permitted (or required) — a policy choice that matters for community colleges and smaller nonprofits without research budget surpluses
• Mandatory travel budget to attend an NSF program meeting in Washington — typically $3,000-$5,000 per team in the Development phase, which must be built into the budget
• No letters of intent or preliminary proposals — teams submit full proposals on the deadline
• Maximum 3 pages for Planning project descriptions; 10 pages for Development project descriptions
• Research.gov or Grants.gov submission

The page limits on project descriptions are unusually tight, reflecting the program's bias toward concrete, demonstrable concept work over extensive theoretical framing. Teams preparing Development proposals should budget significant time for compression rather than expansion.

Translation pathway: I-Corps, Translation to Practice, SBIR/STTR

The solicitation includes explicit language about translational pathways from FINDERS FOUNDRY into NSF's broader innovation portfolio. The most directly named pathways are NSF's Translation to Practice program, Innovation Corps (I-Corps), and the recently relaunched SBIR/STTR programs — meaning that teams completing successful Development awards have a structured opportunity to continue their work as commercialization-oriented research-to-product translation, supported by additional NSF funding outside the FINDERS FOUNDRY envelope.

This is the part of the program design that makes FINDERS FOUNDRY most clearly DARPA-style in orientation. The Foundry model is not designed to produce academic publications as the primary output. It is designed to produce prototype-to-product translations that progress through a defined NSF-supported commercialization pipeline. Teams that approach the program as a traditional education research grant will likely underperform; teams that approach it as a research-to-startup pipeline will likely outperform.

The implication for the small but growing K-12 AI EdTech sector is meaningful. NSF SBIR/STTR has historically been challenging for education companies to win because the program's deep-tech orientation has favored hard-science applications over educational software. A teacher-led, university-validated, parent-co-designed prototype emerging from FINDERS FOUNDRY into the I-Corps and SBIR pipeline has a substantially stronger NSF-internal evaluation case than a cold-start EdTech SBIR application from a company without that research provenance.

Historical context: where this fits in NSF's K-12 portfolio

NSF has been funding K-12 STEM and AI education research for more than two decades through several programs, most notably:

• Discovery Research K-12 (DRK-12) — the primary basic research program for K-12 STEM and computer science education, with awards typically in the $1M-$3M range over 3-5 years
• Innovative Technology Experiences for Students and Teachers (ITEST) — focused on equitable access to STEM career pathways, with awards typically in the $300K-$1.5M range
• Computer Science for All (CSforAll: Researcher Practitioner Partnerships) — research-practice partnerships for K-12 computer science, with awards in the $400K-$1.2M range
• AI Institute and AI Education programs — sector-specific AI investments at varying scales

FINDERS FOUNDRY is not a replacement for these programs. The funding scale ($300K per Development award) is smaller than typical DRK-12 awards, the timeline is shorter (one year vs. three to five years), and the orientation is more prototype-focused than basic-research-focused. It is best understood as a complement to the existing portfolio — a fast, small, applied program that surfaces concepts that could subsequently scale into the larger DRK-12 and ITEST programs, or translate into SBIR/STTR ventures, or both.

The choice to launch this program now, rather than expanding the existing programs proportionally, signals NSF's view that the existing portfolio has been too slow, too academic, and too disconnected from classroom use to deliver on the AI Education EO's stated goals. Whether FINDERS FOUNDRY's structure can actually solve those problems — particularly given the budget pressures NSF is operating under across FY 2026 — will be visible in the November 18 application pool and the Spring 2027 Development award announcements.

What to do before November 18

The Planning deadline has passed. Teams that did not submit Planning proposals in May cannot submit Development proposals in November under the program's structure. For teams that did submit Planning proposals, the operational priorities for the next six months are:

Execute the Planning award deliverables on schedule. The two-month Planning execution window will run through mid-summer. NSF will evaluate Development proposals partly on the quality of the Planning-phase outputs, which means producing real stakeholder engagement data, a real wireframe, and real student pilot results — not paper deliverables.

Solidify the four-stakeholder team. Planning awardees that recruited team members on a tentative basis for the Planning proposal need to formalize Development-phase commitments now, including written collaboration agreements, agreed-upon compensation for non-academic team members, and clear role definitions for the Development work.

Initiate IRB processes early. Development proposals require IRB approval or exemption documentation for the K-12 student involvement that the program mandates. University IRBs typically take 30-90 days for K-12 minor research approvals, and waiting until October to begin the IRB process will create avoidable schedule risk.

Build the translation roadmap. Strong Development proposals will articulate how the prototype work positions for follow-on NSF support through I-Corps, Translation to Practice, or SBIR/STTR. Teams should engage with their NSF program officers during the planning window to validate the translation pathway and identify the right NSF program officers for follow-on engagement.

Plan the showcase deliverable. The Development award includes a Washington showcase event at the end of the funding cycle. Teams that plan a demonstration-ready prototype experience rather than a conference-poster-style presentation will be operating in the orientation NSF has signaled it wants.

For organizations that did not submit Planning proposals, FINDERS FOUNDRY 2026 is closed, but the program's design choices — the staged structure, the mandatory team composition, the translation pathway — are highly likely to recur in subsequent NSF education solicitations and in expanded FINDERS FOUNDRY cohorts in FY 2027 and beyond. The institutional capacity to assemble four-stakeholder teams, run K-12 co-design at IRB-approved depth, and produce prototype-quality deliverables on compressed timelines will be the durable competitive advantage in NSF's K-12 portfolio going forward. Teams that begin building that capacity now will be in stronger positions when the next solicitation arrives.

The Foundry experiment is small. Its implications for how NSF funds K-12 education R&D over the next decade are not.