UW–Madison researcher urges Hudson board to set AI guardrails while exploring classroom uses
Loading...
Summary
Dr. Rich Halverson of the University of Wisconsin–Madison told the Hudson School Board generative AI is transforming teaching and learning and recommended pilot tests, teacher training and policies that preserve student agency and require students to show their work to avoid misuse.
Dr. Rich Halverson, a professor at the University of Wisconsin–Madison, told the Hudson School Board on March 9 that generative artificial intelligence is “the most revolutionary knowledge tool ever built” and urged the district to pursue cautious experimentation and clear policies.
Halverson, who addressed the board after being introduced by the district chair, said AI offers the equivalent of affordable one‑to‑one tutoring and can boost performance, especially for lower‑performing learners, but warned that the technology “is never going to be fully reliable” and will require sustained human judgment. “These systems are never gonna be fully reliable, which means that they're always gonna require people to correct them,” he said.
Board members raised concerns about mental‑health impacts, equity and accuracy. Molly, a board member, told the presenter she worried AI could “steal” creative work and widen gaps between students with home access and those without. Halverson acknowledged those risks and recommended schools provide students with structured AI literacy and require demonstrations of learning — for example, asking a student who submitted a paper to explain key points in person — rather than rely on detection software.
Halverson outlined an “AI‑ready school spectrum” from analog schools that avoid AI to AI‑native models that integrate student co‑design with tools. He recommended two near‑term pathways for districts: one focused on using AI to improve staff efficiency and the other on classroom integration that preserves student agency. He also cited recent studies showing faster, higher‑quality results when users receive some training in AI tools, and warned that versions of these systems will continue to evolve rapidly.
District leaders said the presentation was a thought‑provoker and that the AI task force will continue piloting tools and report back with concrete findings and recommended guardrails over the next several months. No formal policy was adopted at the meeting; board members asked staff to return with results from the task force and specific implementation proposals for further discussion.
