AIR Launches AI in Education Network to Bring Evidence to K-12 AI Use
The American Institutes for Research (AIR) has launched the AI in Education Network-a coordinated set of studies focused on how artificial intelligence is being used across K-12 classrooms, assessments, and district operations. The aim is simple: help schools separate marketing claims from what actually works.
"As AI becomes part of everyday use in education, schools need more than marketing-they need evidence," said Jessica Heppen, President and CEO of AIR. "Our goal is to provide clear, credible insights that help educators, school and district leaders, and policymakers make sound decisions about which types of AI-enabled tools to adopt, and which to avoid."
What the network will study
The network brings together AIR researchers working at the intersection of technology, learning, and evidence. Six studies will examine instruction, assessment, professional learning, and data use-prioritizing practical insights schools can act on.
- Teaching with AI: Testing the Promise and Empowering Educators
Led by Rebecca Bergey
Focus: How professional learning helps teachers use AI responsibly and effectively to strengthen instruction. - 3D Learning With ADAPT-AI: Evaluating Human-AI Collaboration in Science Assessment
Led by Bo Zhu
Focus: Whether human-AI collaboration in science assessment boosts student engagement and performance. - AI in the Classroom: Benchmarking and Testbed Framework for Evidence-Based Adoption in K-12
Led by Ruhan Circi
Focus: A real-world testbed schools can use to evaluate AI tools and make data-informed adoption decisions. - Smarter Data, Stronger Schools: AI-Powered Analytics for K-12 Education
Led by Tammy Kolbe
Focus: Turning district finance and HR data into reliable insights for staffing, budgeting, and strategy. - Fair Play or False Start? Using AI-Generated Data to Examine Validity and Measurement Bias
Led by Sam Rikoon
Focus: Whether AI-generated data can stand in for human responses to test item quality and bias. - Rapid-Cycle Evaluation of CESA 4's LIFToFFS Model
Led by Rachel Chamberlain and Linda Galib
Focus: How AI-enabled tools and educator supports can personalize learning to improve engagement and wellbeing in rural schools.
Why this matters for educators
- Clarity over hype: Expect practical findings on where AI helps (and where it doesn't) across instruction, assessment, and operations.
- Evidence-first adoption: The testbed work supports pilot-and-measure approaches so schools choose tools based on outcomes, not demos.
- Teacher agency: Professional learning is front and center-so AI supports practice rather than dictating it.
- Better use of district data: Finance and HR analytics can inform staffing, budgeting, and resource allocation-if the data are clean and the questions are clear.
- Equity and validity: Studies on AI-generated data and bias will help teams avoid shortcuts that weaken measurement quality.
The AI in Education Network is led by Rachel Garrett, a managing researcher at AIR, who will coordinate learning across the projects and share findings with educators, policymakers, and researchers. "What can AI do well in education, where does it fall short, and how can we make sure it serves students and teachers?" Garrett said. "By linking these projects through a shared learning network, we can move faster, collaborate more effectively, and share insights that help schools and systems make informed decisions."
How to follow and use the findings
AIR will publish updates on a rolling basis through briefs, webinars, and practitioner resources. You can track new releases on the AIR website.
- Use the testbed and study findings to inform your next pilot plan, success metrics, and procurement criteria.
- Align PD with the "Teaching with AI" insights so teachers get hands-on practice with safeguards, prompts, and classroom routines.
- Review data governance before adopting AI analytics-data quality, access controls, and audit trails matter.
- Pressure-test vendor claims with independent evidence from these studies.
Helpful resources
- U.S. Department of Education: AI and the Future of Teaching and Learning (policy guidance and recommendations). Read the report
- Professional learning options for staff exploring AI in instruction and operations: AI courses by job role
Established in 1946, AIR is a nonpartisan, not-for-profit organization conducting behavioral and social science research and delivering technical assistance across education, health, and the workforce. Learn more at air.org.
Your membership also unlocks:
