USD 383 board approves on‑premises AI server, trustees cite data control and instructional uses

The Manhattan‑Ogden USD 383 Board of Education voted to purchase an on‑premises artificial‑intelligence server and three configured models for a quoted cost of $23,000, a measure proponents said will keep district data secure while enabling instructional uses.

Kurt, who moved the item, described the server as a way to provide safeguards while allowing the district to experiment with AI: "this is extremely important for our district ... it's fully customizable, fully secure, and it's private," he said, noting the system would operate without needing Internet access for local model use.

Trustees and staff outlined practical classroom and administrative applications. Administration said the server could process local assessment and MTSS data to identify standards where groups of students need targeted instruction and could host special‑education materials and a searchable handbook as a staff-facing chatbot. "We can take our testing data ... and compare that with state assessments and pull in the state standards," a district staff member explained, adding that built‑in modules can speed common staff queries and save time.

Doctor Allen expressed cautious support, asking the board to keep an eye on tradeoffs and potential efficiencies: "Are there things if this is more efficient than what we're using now that will go away?" he asked, recommending future planning to identify costs and areas of savings.

The board approved the quote 5-0. Administration said web‑blocking and firewall measures will be used to restrict access and to prevent external tools such as ChatGPT from interacting with internal systems.