Get Full Government Meeting Transcripts, Videos, & Alerts Forever!
Get email alerts on the Education Technology topic
No spam. Unsubscribe anytime.
Montana Digital Academy director tells Board of Public Education AI brings promise — and ‘hallucination’ risks
Summary
Jason Neifer, executive director of the Montana Digital Academy, told the Board of Public Education on March 12 that generative artificial intelligence could reshape instruction but also produces convincing false information and images unless teachers and systems prepare for it.
Get email alerts on the Education Technology topic
No spam. Unsubscribe anytime.
Jason Neifer, executive director of the Montana Digital Academy, told the Montana Board of Public Education on March 12 that generative artificial intelligence is already producing realistic-seeming classroom images and written materials — and that schools need plans to use the tools responsibly.
"What I'm hoping that you get before the end of our time today is the opportunity to see that artificial intelligence is growing very quickly, and more importantly, that there are both opportunities and potentials for peril," Neifer said as he opened a 30‑minute demonstration for the board.
Neifer showed examples from image-generation tools (MidJourney and others) and from large language models, including a Google Gemini demo, to illustrate how brief natural‑language prompts can yield images and essays that many viewers would treat as authoritative glance media. He demonstrated that adding camera and photographic‑style terms to prompts can make generated images look like real photographs, and he cautioned that faces and details can still be wrong even when the image looks realistic.
"This image is generated by an artificial intelligence tool that's called MidJourney," Neifer said, and later described the widely observed problem that some models invent facts or biographies: "That's wrong, as it turns out… That's called hallucinating. I don't like that term, but I like to refer to it as 'wise.'" He urged caution when treating model output as verified information.
Neifer also walked through classroom applications: converting content into age‑appropriate formats, creating glossaries and vocabulary suited to a 7th‑grade reader, and generating differentiated instructional materials. He argued that the technology could help teachers personalize learning and assist instructional designers, special‑education teachers and librarians with new tools, while noting cost and equity limits — simulated demos often showed devices and equipment schools cannot afford.
Board members asked about legislative timing and funding for a proposed 'Frontier Learning Lab' at the Montana Digital Academy. Neifer said the academy had already pitched the proposal and that it had made it through an initial appropriations step; he told the board he hoped for a clearer answer by mid to late April if the appropriation proceeds.
Neifer recommended that educators be trained both in using AI productively and in teaching media literacy so students learn to evaluate machine-generated content. "We need to help teachers first understand that this is not just a vector for cheating," he said, arguing training and governance are necessary if schools adopt these tools.
The board did not take formal action on the presentation; the session moved on to other agenda items after a brief question-and-answer period.

