Get Full Government Meeting Transcripts, Videos, & Alerts Forever!
Get email alerts on the Ai Policy topic
No spam. Unsubscribe anytime.
Red Hook trustees begin discussion on draft AI policy, ask for more review and limits on use in interpretive drafting
Summary
Trustees discussed two draft AI policies and asked staff and trustees to refine rules that would allow AI for factual research while restricting or flagging its use for interpretive drafting (e.g., local laws, legal memoranda), and requested clearer disclosure language for any AI‑assisted content.
At the March meeting the Red Hook Village Board began a preliminary discussion about adopting an AI policy for village staff and trustees, with a focus on disclosure and limits for interpretive or legal drafting.
Mayor Karen Smith presented two draft policies for discussion—one modeled on another municipality’s recently adopted policy—and asked trustees for initial feedback. Trustee Fran said she had used AI to pull supplemental historic information for the proposed historic‑resources table and supported requiring a note explaining how the table was prepared; that note was incorporated into the motion concerning the table.
Several trustees urged caution. Trustee Amy, who has experience as a public librarian, said factual lookups are not the same as interpretive drafting. “Researching policy options, drafting correspondence, memoranda, local laws, resolutions—those are interpretive purposes,” she told colleagues, arguing such tasks require human professional judgment and should be treated differently in policy. Another trustee urged that any AI use be accompanied by verifiable source citations and, where appropriate, the specific proprietary tool used.
The mayor said the board would not adopt a policy at the meeting; trustees voted to defer action and asked staff to return with a refined draft that distinguishes factual research assistance from tasks that involve legal or policy interpretation.
What’s next: staff will refine the draft AI policy language, with attention to (1) required disclosure language and attribution for AI assistance, (2) allowable uses for staff vs. elected officials, and (3) safeguards for legal and planning documents.
