AI-generated summary, not yet reviewed by Better Utah staff. Please consult the bill text.

Utah law currently requires police departments to have written policies governing their use of generative artificial intelligence—tools like ChatGPT that can draft text, images, or other content based on learned patterns. This bill builds on that requirement by adding new definitions, including a concept of "oversight or safeguard settings" (such as required human editing, AI-generated content disclaimers, or limits on which crimes AI tools can be used for), and requires agencies to disclose in their policies which AI technologies officers may use, what safeguard settings are available for each, and whether the agency requires, disables, or allows officers to bypass those settings; this disclosure requirement would not apply to administrative tools like spell-checkers or productivity software that don't materially affect investigations. Police reports created with AI assistance must still include a disclaimer noting AI involvement and a certification from the officer that they reviewed it for accuracy, and agencies must review and update their AI policies annually. The bill also requires agencies to publicly post their AI policies on their websites (or on the state's public notice website if they lack one), updating postings within 30 days of any changes, though this posting requirement would not apply to prison or jail security and safety operations. Failure to comply with these requirements would not prevent an agency from using AI technology or from using evidence created with it.