At the Marysville Exempted Village School District board meeting on Nov. 20, district instructional leaders described a phased rollout of generative artificial intelligence tools, emphasizing teacher training, data protections and community review before students will use the systems. Chris Campbell, the district’s director of teaching and learning, said pilots will start in two buildings and that the district has blocked student access to public models such as ChatGPT until guardrails are in place.

The presentation framed generative AI as a way to advance the district’s personalized‑learning strategy by automating routine tasks, producing differentiated instructional materials and providing targeted academic and behavioral supports. “It’s going to help teachers adapt content and create pathways for each student,” Campbell said, describing pilot work with vendors including SchoolAI and PowerBuddy and a design partnership to integrate AI with Schoology and PowerSchool.

Mark Gallagher, director of continuous improvement, defined generative AI as systems that create original content from natural‑language prompts and cautioned about bias and “hallucination” (false outputs) in model responses. The district reported internal surveys of roughly 100 staff and more than 475 students; Campbell said nearly 80% of staff use some form of generative AI and about 90% believe AI will affect education in the next five years, while student sentiment was mixed (about 35% very unfavorable, about 34% favorable).

Board members asked pointed questions about limiting student access and grade‑level differentiation, verifying sources and aligning tools to district curriculum. Campbell said the district will design constrained instructional “spaces” (prompted environments) that pull only from district‑approved resources and use tutorial‑style prompts that nudge students instead of simply giving answers. “We’re not letting ChatGPT run wild in the district,” he said, adding that IT has locked down unvetted access until pilots and communications are complete.

Gallagher offered a specific use case on behavioral supports: after professional learning, an AI tool could generate a tailored plan for reinforcing desirable behavior, suggest de‑escalation techniques and even draft phone‑scripts for difficult family conversations to help new teachers practice crucial interactions.

Public commenters urged the district to require fact checking and guard against bias; one asked whether AI accuracy (which he characterized as sometimes low) and fairness are being addressed. Campbell and Gallagher acknowledged those concerns and said the steering committee and upcoming community sessions will focus on guardrails, transparency and how the district will evaluate vendor assurances about data protection.

Next steps include continuing pilots through December, returning to teacher and student committees in January to draft guardrails, a public review of policy and guardrails with community stakeholders in March, and board consideration of a formal AI policy in June — ahead of an Ohio requirement the presenters said applies to districts next summer. The district said it will circulate staff and family communications about the pilots and invite stakeholders to review classroom use and guardrails at a March engagement session.