Get Full Government Meeting Transcripts, Videos, & Alerts Forever!
Get email alerts on the Technology Policy topic
No spam. Unsubscribe anytime.
Eureka Council adopts AI policy after lengthy debate; staff to return with revisions
Summary
After an extended council discussion and public comment, the council adopted Policy 1.14 limiting staff use of generative AI, requiring transparency when AI produces original content, forbidding input of sensitive data to external tools, and directing staff to return with revisions by the council's second April meeting.
Get email alerts on the Technology Policy topic
No spam. Unsubscribe anytime.
Eureka’s City Council voted 4‑0 on March 17 to adopt an artificial intelligence policy (Policy 1.14) that sets citywide rules for using generative and limited‑memory AI tools. The adoption followed more than an hour of debate about definitions, ethical concerns, operational risk and the pace of AI adoption.
Finance Director Malar, who presented the draft policy, said its twin goals are data protection and proper usage. The policy as adopted instructs staff not to submit sensitive personal or proprietary information into public AI systems, to disclose when AI has ‘‘significantly contributed’’ to written content and to verify factual claims generated by AI. Malar stressed the need to avoid treating AI outputs as raw, authoritative research and urged staff to locate primary sources when relying on AI summaries.
Several council members voiced strong reservations. Council member Moulton said she opposed city use of AI in its current form and argued for very tight limits: "I don't think the city should use it at all," she said during debate, while acknowledging the technology is already in use. Council member Castellano asked for clearer definitions distinguishing formatting tasks (simple editing) from structural, consequential uses (policy drafts or legal analysis). Council member Bauer urged caution and a more thorough review before broad adoption; he moved to table the item but the substitute motion failed. After discussion, council moved to adopt the policy with a friendly amendment directing staff to refine definitions and return with a revised draft no later than the council’s second meeting in April.
Council also formed an ad hoc working group (Council members Moulton and Castellano) to meet with staff on the policy revisions. Director Malar suggested the city could also convene a small working group or study session to isolate risk areas and advise where enterprise software that already embeds AI should be treated differently than external tools.
The policy adoption is an administrative action (policy rather than ordinance); the council voted to adopt the policy and to require a revised draft back for council review by the second meeting in April.

