Montana Digital Academy urges 'safe, responsible, specific' AI use in schools and pilots for teachers

Jason Neifer, executive director of the Montana Digital Academy, told the task force on Feb. 12 that schools should treat modern AI as "augmented intelligence" and focus on safe, responsible and specific applications rather than adopting technology for its own sake.

Neifer described MTDA programs (distance learning, EdReady Montana, the Digital Learning Clearinghouse) and the Frontier Learning Lab pilots: an AI help desk for teachers, professional development, curated secure AI platform access for K–12, and experiments in tutoring and data‑driven decision making. He warned of common pitfalls — hallucination, "AI sycophancy" (models favoring agreeable responses), bias and student‑privacy risks — and emphasized adult training (teachers, administrators, librarians, parents) before expanding student use.

On procurement and policy, Neifer recommended clearer student‑data privacy rules aligned to existing Montana law (MOPIPA) and suggested the state create funding and procurement pathways to give schools vetted, secure AI tools at scale. He urged experimentation with targeted pilots (for tutoring, assessment and adaptive learning) and cautioned against turning AI into a turnkey "vending‑machine" solution that does the thinking for students.

Members asked about hallucination rates, prompting and model configuration; Neifer recommended retrieval‑augmented generation (RAG) and careful prompting for lower hallucination and shared examples of teacher workflows where AI acts as a thought partner rather than a replacement. Several members invited MTDA to join workgroups and contribute to the July 1 report.

No formal policy was adopted at the session; chairs asked MTDA to share materials and participate in further meetings and workgroups.