Almost half of the largest U.S. school districts have not adopted any formal policy on the use of artificial intelligence, according to a landscape scan published by the Center for Education Market Dynamics. The scan examined publicly available documents from 219 districts that together serve more than 14.6 million students, roughly 29 percent of all public school students, across 43 states. It found that 44 percent of the districts had no AI policy adopted or in progress as of the start of the 2026-27 school year.
Among the districts that did have policies, only 10 percent of the policy documents referenced professional learning or training for staff. The scan found that existing policies were often scattered across multiple documents, webpages, or frameworks rather than forming a coherent single policy, and few were confirmed to have been developed with community involvement. The tone of most policies emphasized restrictions over structured guidance for productive integration.
The scan detected no statistically significant relationship between the presence of an AI policy and district characteristics such as size, demographics, urban or rural location, or curriculum selection. The finding suggests that policy adoption so far has not tracked with the structural factors that often shape education technology decisions.
States shift from guidance to mandates
While most states continue to rely on voluntary guidance or no framework, a small but growing number have enacted laws requiring districts to adopt AI policies. Ohio’s revised code section 3301.24, effective September 2025, required the state to develop a model policy and obligated every school district, community school, and STEM school to adopt its own policy by July 1, 2026. Maryland’s SB 720 (the Artificial Intelligence Ready Schools Act), effective June 2026, requires each local school system to adopt an AI policy aligned with state guidance, designate a central-office AI coordinator, and use a state-developed rubric when evaluating AI tools for procurement.
Virginia’s HB 1186 and SB 394, taking effect July 2026, require the state to issue guidance covering student data privacy under FERPA and COPPA, teacher training, transparency, and data-processing agreements that prohibit using student data to train external AI models, according to a compilation by TrueMadeAI. The statute requires school boards to establish policies consistent with that guidance but sets no calendar deadline. Utah’s HB 273, effective July 2026, requires the state board to publish model balanced-technology and AI-use policies by December 2026, with local education agencies required to adopt policies by July 2027. Utah’s law mandates parent notification for classroom generative AI use, prohibits biometric surveillance and psychological profiling, requires LEA approval of student-facing AI tools, and mandates staff training, with the possibility of withholding digital literacy funds for noncompliance.
Mandates versus guidance
A separate analysis of 122 districts and schools across 38 states by EdSurge and the AI and Education Studio found that only 15.6 percent explicitly referenced state guidance in their own policy documents, despite 63 percent being in states with officially published AI guidance. The analysis concluded that mandatory state guidance drives district adoption more reliably than advisory frameworks and that states relying on voluntary guidance saw inconsistent local adoption. In contrast, states such as Massachusetts, whose 2026 guidance according to TrueMadeAI explicitly leaves implementation decisions to local leaders, provide frameworks without requiring district adoption.
Research evidence points to slow and uneven adoption
Earlier peer-reviewed research found an even lower baseline. A content analysis of 9,229 U.S. K-12 districts published by Researcher.Life and Shaping Education Policy found that as of May 2024 only 14.13 percent of the districts that had identifiable policies had an AI policy. The study characterized the results as documenting the initial phase of AI policy formation in U.S. schools.
A survey by the Center on Reinventing Public Education (CRPE) of state education agencies found that most states had not developed coherent AI approaches. Among 28 surveyed states, about half had updated standards for AI literacy, but fewer than half had issued mandated policies for districts or assisted with procurement.
