Skip to content

Reports

Beyond Adoption: A National R&D Agenda for Generative AI in K–12 Mathematics and Science Education

This white paper synthesizes findings from a cross-institutional faculty collaboratory across 13 higher education institutions, examining how educator preparation programs (EPPs) can intentionally integrate generative AI into teacher education coursework. Rather than treating generative AI purely as a tool adoption challenge, the authors frame integration as an instructional design problem where AI serves to scaffold and support professional judgment rather than bypass candidates’ pedagogical reasoning. Drawing on 11 documented course implementations spanning methods, assessment, linguistics, and clinical practice, the paper outlines practical instructional models, key implementation lessons, and actionable guidance for faculty, program leaders, and accreditation stakeholders seeking to cultivate critical AI literacy and responsible classroom practice.

Open

Adopted but Unstructured: Findings of GenAI Use, Training, and Policy in K-12 Math and Science Education from a 2026 National Teacher Survey

This report presents findings from a national survey fielded through RAND’s American Teacher Panel examining how K-12 public school math and science educators adopt and use generative AI, tracking year-over-year shifts in classroom practices, professional learning, and district policies. Teachers primarily use GenAI for instructional planning, content creation, and differentiating learning materials, with overall adoption rising to 58% alongside notable variations linked to teachers’ prior experience and usage intensity. Descriptive analyses explore teachers’ perceptions of student learning, instructional time allocation, and persistent barriers to implementation, highlighting substantial gaps in formal training and district guidance to provide nationally representative evidence for policymakers and educational leaders navigating GenAI integration.

Supplement Documents:
Survey InstrumentTechnical Document

Open

AI Integration as Instructional Design: Lessons from a Cross-Institutional Faculty Collaboratory in Teacher Preparation

This research agenda is informed by two years of representative survey data from American Teacher Panel, establishing priority directions for generative AI in K-12 mathematics and science education. While adoption has expanded rapidly among educators, usage remains predominantly teacher-facing and centered on instructional planning, revealing critical structural gaps in guided student learning, assessment evaluation, professional development, and district policy. Moving beyond perception-based surveys, the agenda establishes six core priority areas and five cross-cutting commitments—emphasizing objective classroom artifacts, observational metrics, and domain-specific tool design—to mobilize researchers, developers, district leaders, and funders in building actionable, rigorous evidence ahead of entrenched educational practice.

Open

Generative AI in K-12 Classrooms: A Mid-year Implementation Report Across School Districts

This report presents mid-year findings from a year-long study with public school districts examining how teachers adopt and use Colleague AI, with platform usage linked to district administrative records on teacher and student demographics, attendance, and test scores. Teachers primarily use AI for lesson planning, standards alignment, and adapting classroom activities, with usage patterns varying by years of experience, credential status, and the demographic composition of their classrooms. Regression analyses explore correlational associations between teacher AI usage and changes in student interim assessment scores, contributing early quantitative evidence to inform future causal studies of generative AI’s impact on K-12 student outcomes.

Open

More Reports

Accessible Accordion

This report from the AmplifyGAIN Center explores how school policy, social context, and pedagogical orientation shape K-12 STEM teachers’ adoption of generative AI. Drawing on an exploratory study, the authors examine whether teachers use AI for routine substitution or transformative student learning. Findings indicate that while many use cases focus on efficiency, student-centered beliefs and administrative backing catalyze meaningful classroom innovation. The report provides clear insights for education leaders aiming to align institutional guidelines, professional development, and resources to improve instruction.

Open

This WestEd Perspectives brief draws on interviews with more than 60 educators to surface what teachers actually want from GenAI professional learning: sessions that connect to existing instructional priorities, are hands-on and grounded in classroom practice, meet a range of readiness levels, and unfold over time through collaboration. The brief translates these insights into concrete actions for state and district leaders.

Open

This report presents findings from a 7-week co-design pilot with 21 Washington State teachers and 600+ students showing that AI tools function most effectively as a “third agent” in classrooms when teachers provide clear scaffolding and framing, with narrative feedback valued over numeric scores and Student Growth Insights enabling real-time instructional adjustments.

Executive Summary Full Report

This report presents findings from a nationally representative survey of US public school math and science teachers examining their generative AI adoption, classroom use, perceptions of student learning impacts, and institutional support needs as educators navigate rapidly evolving AI integration pressures.

Open

Supplement Documents:

Survey Instrument

Technical Document