Get Full Government Meeting Transcripts, Videos, & Alerts Forever!
Get email alerts on the AI Guidelines topic
No spam. Unsubscribe anytime.
POST presents draft AI guidelines for law enforcement, says human review and oversight must remain central
Summary
POST staff presented a draft statewide guideline addressing AI uses across report writing, dispatch, investigations and records; workshops will continue and staff emphasized human review, disclosure, training and vendor accountability as core requirements.
Get email alerts on the AI Guidelines topic
No spam. Unsubscribe anytime.
Mark Nagel, a law-enforcement consultant with POST, told the advisory committee that artificial intelligence is already entering police operations — in report drafting, transcription, dispatch, investigations and records review — and that a statewide professional guideline is needed to prevent inconsistent practices and legal exposure.
"AI may assist law enforcement personnel, but it cannot replace human judgment, legal responsibility, or professional accountability," Nagel said, laying out a framework that emphasizes verification, disclosure of AI use, retention of AI-origin drafts and audit trails, supervisor oversight, and training requirements. He said the guideline builds on the statutory floor created by SB524 (which requires disclosure and retention practices for AI-generated report drafts) and expands guidance to procurement, vendor accountability, records management and supervisory practices.
Nagel described the workshop process supporting guideline development. POST held a subject-matter expert workshop on April 15 with law enforcement leaders, prosecutors, defense representatives, legal scholars and technologists to pressure-test the draft; a second workshop was scheduled for June 10 to finalize refinements before internal review and possible commission consideration.
Committee members pressed on implementation details for smaller agencies and how dispatch/records applications would be governed. Nagel said the guideline includes a dedicated chapter on training and is intended to govern AI systems broadly rather than single vendors, with the consistent theme of human verification and documentation.
What happens next: POST will hold the second workshop, refine the draft, and prepare the guideline for internal review and prospective commission consideration; the advisory committee will receive updates as the process continues.

