Utah official details statewide AI training, tools and policy to BESE committee

Matt Winters, AI education specialist at the Utah State Board of Education, told a meeting of the State Board of Elementary and Secondary Education committee that Utah has used long-term investment, coordinated procurement and broad collaboration to scale artificial intelligence in K–12 classrooms.

“we were lucky enough to have a half million dollar grant here in the state of Utah to go out and work with teachers across the state to prepare them for artificial intelligence integration,” Winters said, describing a three-tier professional development program that he said has reached about 7,000 teachers and supported roughly 1,500 teachers through advanced coursework and lesson-plan development.

Winters stressed that the state paired training with purchasing and policy. He described an April 2024 statewide RFP run through Utah Education Network that gave districts access to consortium-priced AI tools (named in the presentation as SchoolAI, Magic School and Skillstruck Chat for Schools; UEN later added Brisk). “Our RFP allows any of our districts or charter schools across the state to buy 3 AI tools … at a low cost,” he said, adding that consortium purchasing reduced per-district price and drove wider adoption.

Why it matters: Winters presented procurement, PD and a research agenda together as a package to reduce barriers to classroom implementation while creating shared expectations for data privacy and teacher practice. He said the state’s AI framework — adopted in April 2024 and described in the presentation as a “living document” — guides district policy and professional development.

Key details and claims Winters provided specific counts and program details: the Utah Digital Teaching and Learning Grant is a legislatively funded program with annual appropriations he said typically range from about $18 million to $22 million; a $50 million legislative set-aside funded replication of a model school program (the Catalyst Center); and teacher PD reached “about 7,000” educators for under $1 million in grant funding. Winters said teachers’ self-reported measures of AI use and comfort improved “about 1 to 2 standard deviations” in preliminary pre/post analyses.

On policy and regulation, Winters described Utah’s Office of AI Policy, created by the legislature and standing up in mid-2024, as having a regulatory role. “Doctor Boyd… has a regulatory hammer,” Winters said, explaining that the office can pause or recommend code changes and that Utah has enacted laws addressing some AI applications — for example, a state code pertaining to mental-health chatbots that Winters said the board is examining for K–12 applicability.

Winters also highlighted statewide practices intended to protect student information. He described a required data-privacy agreement format (an NDPA version 2 with Exhibit E) that districts can sign and share across the consortium, and said UEN performs data-privacy checks and holds contracts. If a local district lacks a signed DPA for a tool, Winters said the tool’s classroom use would be a contractual breach.

Questions from the committee and others focused on funding sources, metrics and local control. Winters said the PD funding referenced was provided by Intermountain Healthcare and that UEN holds detailed contract-usage metrics; he acknowledged variance between vendor-reported adoption figures and district-reported measures. On measuring academic impact, he described a fresh $200,000 pilot (15–20 administrative teams) that will include pre/post assessment comparisons and parent engagement components, with results expected in roughly a year.

Winters repeatedly emphasized voluntary, collaborative alignment rather than top-down mandates. He said Utah’s approach combined state-funded programs, local control for districts to choose tools, and sector-wide collaboration including higher education, industry and the University of Utah’s Responsible AI Initiative.

What the presentation did not resolve: Several claims Winters cited (percent adoption rates, measured gains in teacher practice and the effectiveness of specific tools) were presented as program-reported or preliminary survey results; the committee asked for follow-up and deeper, longitudinal evidence tying AI-enabled practices to student outcomes. Winters said research collaborations and pilot projects are in progress.

Next steps reported to the committee included publishing lesson plans and PD materials on state web portals and following up on the pilot study and additional competency frameworks Winters said would be released in coming months.