Council approves generative-AI use policy requiring human review before publication

Tremonton's council approved a generative-AI policy that provides guidance to city employees, contractors and council members on acceptable AI use. The policy includes explicit prohibitions on placing private or protected information into AI prompts and requires a human reviewer to validate outputs before materials are distributed externally.

"Nothing can go out of AI unless it's been read or touched by the person that is working on it," the Mayor explained when summarizing the policy's "human touch" requirement. Staff said the policy has been vetted against Utah state code and will be finalized by the city attorney and procurement staff prior to an August 15 effective date. Council voted to include the policy in the consent agenda for adoption and approved it as part of the consent items.

Council members and staff emphasized that the policy is intended to enable efficient use of AI for public-document review, ordinance templates and research while preserving accountability and data protections. The city will publish the employee handbook and provide department-level training on the new rules.