View in EDS

K-12 AI Infrastructure: Findings from Educator and Developer Outreach. K-12 AI Infrastructure Program

Saved in:
Bibliographic Details
Title: K-12 AI Infrastructure: Findings from Educator and Developer Outreach. K-12 AI Infrastructure Program
Language: English
Authors: Rebecca Griffiths, Vanessa Peters Hinton, Shelton Daal, Megan Pattenhouse, Digital Promise
Source: Digital Promise. 2026.
Availability: Digital Promise. 1001 Connecticut Avenue NW Suite 935, Washington DC 20036. Tel: 202-450-3675; e-mail: contact@digitalpromise.org; Web site: https://digitalpromise.org/
Peer Reviewed: N
Page Count: 30
Publication Date: 2026
Document Type: Reports - Research
Education Level: Elementary Secondary Education
Descriptors: Elementary Secondary Education, Artificial Intelligence, Technology Uses in Education, Educational Technology, Needs Assessment, Educational Benefits, Barriers, Program Evaluation, Formative Evaluation, Data, Information Security, Privacy, Curriculum, Bilingual Students, Students with Disabilities, Rural Areas, Faculty Workload, Well Being, Safety, Evaluation Criteria, Benchmarking
Abstract: This report presents findings from the K-12 AI Infrastructure Program's outreach to educators, edtech developers, and other stakeholders, conducted during late 2025 to mid-2026. It examines what educators want from AI tools, where current AI applications fall short, how education technology developers are responding, and what public infrastructure is needed to bridge these gaps. Educators want AI systems grounded in their district's curricula, values, and student populations, capable of supporting formative assessment at scale, and able to return time to overburdened staff. However, generic AI tools can undermine curriculum coherence, and student-facing applications raise concerns about effectiveness, wellbeing, and safety. Existing trust proxies, such as certifications, fail to address underlying privacy and fairness issues, while evaluation standards for AI-based edtech remain immature. The report identifies five priority infrastructure gaps: knowledge graphs supporting ambient assessment, child speech recognition and representative voice datasets, longitudinal and holistic student data systems, shared evaluation benchmarks, and context engineering infrastructure. The authors argue that publicly available, modular infrastructure--rather than proprietary, vendor-specific solutions--is needed to help AI tools better align with pedagogical goals while protecting student privacy and serving historically underserved populations.
Abstractor: As Provided
Entry Date: 2026
Accession Number: ED682349
Database: ERIC
Description
Abstract:This report presents findings from the K-12 AI Infrastructure Program's outreach to educators, edtech developers, and other stakeholders, conducted during late 2025 to mid-2026. It examines what educators want from AI tools, where current AI applications fall short, how education technology developers are responding, and what public infrastructure is needed to bridge these gaps. Educators want AI systems grounded in their district's curricula, values, and student populations, capable of supporting formative assessment at scale, and able to return time to overburdened staff. However, generic AI tools can undermine curriculum coherence, and student-facing applications raise concerns about effectiveness, wellbeing, and safety. Existing trust proxies, such as certifications, fail to address underlying privacy and fairness issues, while evaluation standards for AI-based edtech remain immature. The report identifies five priority infrastructure gaps: knowledge graphs supporting ambient assessment, child speech recognition and representative voice datasets, longitudinal and holistic student data systems, shared evaluation benchmarks, and context engineering infrastructure. The authors argue that publicly available, modular infrastructure--rather than proprietary, vendor-specific solutions--is needed to help AI tools better align with pedagogical goals while protecting student privacy and serving historically underserved populations.