Utah County adopts concise AI acceptable-use policy to protect resident data

Utah County commissioners on April 10 adopted a concise artificial intelligence acceptable-use policy intended to allow staff to benefit from AI tools while protecting resident data and limiting county liability.

Patrick Worrow, director of the county’s Information Systems Department, said an advisory committee spanning legal, IT, HR, the auditor’s office and several departments drafted the two-page document. The policy identifies prohibited uses (including inputting personally identifiable information), places accountability on staff to verify AI outputs, and contemplates required training before employees use AI in county work.

Worrow said the county weighed risks including inadvertent disclosure of private data, intellectual-property infringement, and the tendency for AI outputs to appear authoritative when they may be incorrect. "A human has to be responsible," he told commissioners, summarizing the policy's requirement that staff proofread and take ownership of any AI-generated product.

Commissioners praised the document’s brevity and cross-departmental input and voted to adopt the policy by voice vote. The adopted approach allows controlled use of AI tools while directing staff training and continued review of the policy as technology evolves.

Next steps: the advisory committee will roll out training for employees who want to use AI tools and will update the policy as use cases and risks evolve.