Get Full Government Meeting Transcripts, Videos, & Alerts Forever!
Get email alerts on the Education Technology topic
No spam. Unsubscribe anytime.
Moore County Schools outlines phased generative AI guidance and training for 2025–26
Summary
Moore County Schools on Nov. 3 outlined district guidance and a phased rollout for generative AI, restricting direct student chatbot use in K–5, permitting supervised use in grades 6–8, and allowing guided student use in grades 9–12; staff emphasized training, disclosure, and safety controls.
Moore County Schools presented a detailed approach to generative artificial intelligence for the 2025–26 school year, emphasizing teacher training, student safety, and a phased rollout of tools and permissions.
Assistant director for technology Will Herring told the board the district set three "red lines": do not allow AI to undermine teaching and learning, do not expose students to unsafe products, and do not expose children to tools before they are developmentally appropriate. Herring said the district intentionally limited student access to two Google‑integrated products this year — Gemini (Google’s generative model) and NotebookLM — so staff could manage risk and maintain visibility through Google EDU account controls.
Herring walked trustees through the phased access plan: K–5 students will not engage directly with chatbots; teachers may use AI tools in whole‑class demonstration and explicit "think‑aloud" instruction. Grades 6–8 may use bots with direct supervision. Grades 9–12 will have guided, teacher‑directed use for drafting and polishing work. Herring emphasized the district default for assignments is level 0 (no AI) unless the teacher explicitly assigns a higher permitted level, and that students must disclose use and provide chat logs or prompts when they rely on AI in a graded assignment.
Staff reported training and rollout numbers: 655 teachers completed district training and approximately 7,507 students in grades 6–12 have viewed and acknowledged a short student training video. Herring described teacher training as aimed at demystifying the tools, teaching safe prompt use, and integrating AI in ways that support — not replace — existing instructional frameworks.
Herring highlighted two product choices for particular reasons: Gemini offers creative brainstorming and prototyping, while NotebookLM can be limited to preloaded, vetted source materials the user chooses. "Part of our Google EDU account... Google has agreed to not train against our students, our staff, anything that gets input into that," Herring said, noting the district’s preference for tools that minimize the risk of student data being used to train public models.
Board members raised practical questions about teacher workload, how teachers will coach students to use permitted AI levels, the district’s ability to monitor prompts, and behavioral risks such as students forming unhealthy attachments to chatbots. Staff said prompt records can be produced (for example by copying prompt text into Google Docs) and that NotebookLM instances are built around user‑fed content rather than the open internet.
Herring said the district will continue iterative training and will convene teacher cohorts and technical leaders to refine coaching strategies so classroom teachers can reasonably manage allowed AI uses without a disproportionate policing burden. Herring and other staff framed the district’s approach as cautious and iterative rather than an all‑in or all‑out mandate.
The board thanked staff for the presentation and asked staff to continue rapid training and to report back on how the policy operates in practice as the school year progresses.
