In response to the call for feedback by the U.S. Department of Education relating to their grantmaking proposal on artificial intelligence (AI) and education, Accelerate has filed a comment highlighting our recent Call for Effective Technology (CET). AI was highlighted by U.S. Education Secretary Linda McMahon in her “Dear Colleague” letter that listed “AI-Enhanced High-Impact Tutoring” among her key priorities. The opportunity to file this comment was well timed as Accelerate just completed reviewing 160 applications and finalizing grant awards to study 10 AI-powered K-12 tools through the Call for Effective Technology.
Priority on Advancing Artificial Intelligence in Education
Submitted via www.Regulations.gov
August 2025
Submitted by: Accelerate – The National Collaborative for Accelerated Learning
Accelerate appreciates the opportunity to submit this Comment in response to the U.S. Department of Education’s proposed supplemental priority on Advancing Artificial Intelligence in Education. We commend the Department’s recognition of AI’s growing relevance in instructional improvement, personalized learning, and educator support. We support the proposed framework’s alignment with evidence, innovation and personalized learning.
As a national nonprofit focused on high-impact, evidence-based solutions in education, Accelerate recently administered the Call for Effective Technology (CET) program—a competitive grant and evaluation initiative to identify and support promising AI-powered and ed-tech tools that improve instructional effectiveness in real, public, K-12 classrooms. Aligned with our comprehensive and targeted eligibility criteria, the program attracted 160 applications, which we narrowed to our final cohort of 10 grantees (see Appendix).
Below, we offer observations and recommendations grounded in this experience that we hope will inform and strengthen the Department’s proposed Priority (b).
1. Real-World Implementation Should Be a Precondition for AI Readiness
CET prioritized tools ready for deployment in PK–12 public schools by requiring working prototypes, district partnerships, implementation in at least two districts, and evidence of at least initial pilot testing with its target users. We recommend the Department incorporate similar readiness standards to ensure selected tools are beyond the conceptual stage, have demonstrated basic functionality with their target users, and suitable for field evaluation.
2. Build the Evidence Base Through Embedded, Practical Evaluation
All CET grantees are required to participate in structured evaluations conducted by an external research partner. This research focuses on usability, feasibility, and early evidence of learning impact. We encourage the Department to make research participation conducted by a third party in live classroom settings a core component of any AI-related funding competition and to emphasize evaluation designs that are practical, policy-relevant, and timely.
3. Center Access and Affordability in Tool Design and Evaluation
CET’s research protocols include tracking to help understand who is accessing AI so we can evaluate reach and affordability over time. At a moment when many students are below grade level, in need of remedial or developmental education, or otherwise in need of additional assistance, we urge the Department to incorporate similar criteria focused on increasing access into its program priorities and reporting structures.
4. Recognize Educator Engagement and Feedback as Core to AI Integration
CET tools are required to leverage and loop-in teacher expertise by design and collect educator feedback as part of the evaluation process. We recommend the Department prioritize applications that engage educators as co-designers and active users of AI systems, not passive recipients.
5. Fund Comparative Research on AI-Enabled Tutoring and Instructional Models
Several CET awardees offer AI-powered or hybrid tutoring platforms. Our research will begin to compare AI-only, hybrid, and human-led models to assess cost, usage, engagement, and student outcomes. We recommend the Department invest in such comparative analyses, particularly in high-impact tutoring.
6. Adopt a Clear, Research‑Aligned Definition of “High‑Impact Tutoring”
Tutoring or “High-Impact Tutoring” is mentioned in the Proposed Priority (b) items (iv), (v) and (ix) – and we propose the following definition: High‑impact tutoring is individualized, small‑group instruction (student‑to‑tutor ratio ≤4:1) delivered by a consistent tutor (in-person tutor, synchronous virtual tutor, or evidence-based tech-enabled tutoring program) during the school day, three to five times per week for 15–60 minutes, and sustained to approximately 50 hours across the school year. It uses high‑quality instructional materials aligned to grade‑level Tier 1 content and incorporates direct instruction with frequent formative assessment. Tutoring functions as a tiered support that reinforces core instruction, is not homework help, and should be scheduled so students do not routinely miss non‑core classes (e.g., arts/electives).
7. Incentivize Cost Transparency and Scalability Planning
CET requires detailed costing data to assess feasibility and long-term value. Detailed cost data on program implementation is necessary to assess the cost faced by schools and the return-on-investment (i.e., cost-effectiveness) of an educational intervention such as high-impact tutoring. The Department could adopt similar standards to support responsible procurement and avoid hidden costs in scaling.
8. Structure AI Grant Program Execution via Qualified Intermediaries
While we support the Department’s proposed priority, we strongly recommend that any resulting funding program be implemented through qualified intermediaries rather than directly by the Department. Vetting AI-powered tools for classroom use requires nuanced review of instructional alignment, data security practices, technical feasibility, and district readiness—tasks that demand deep content expertise, contextual knowledge, and dedicated capacity.
Intermediaries such as Accelerate are better positioned to conduct this work through multi-phase application reviews, structured onboarding, field-based evaluation, and district-level implementation support. This structure not only ensures a higher quality of tool selection and support but also promotes nimble research cycles and field-informed scaling. We urge the Department to prioritize this model in program design and execution.
Conclusion:
The Accelerate CET program affirms that it is possible to responsibly pilot AI in schools, augment on-going instruction, increase educator capacity, and build an evidence base without slowing innovation. We appreciate the Department’s leadership and welcome opportunities to share data, insights, and tools from our portfolio to help advance this national effort.
