The City Council approved new rules for artificial intelligence on October 8, setting limits on how city staff use it and what residents must be told when it plays a role in city business. The vote was 5–0, KESQ reported.
Under the policy, AI communicating directly with the public for the city must identify itself, offer a practical way to reach city staff and meet accessibility requirements.
When a high-risk use materially influences a decision about an identifiable person, the city must tell that person AI played a material role and provide a practical way to seek human review or correction. Those uses can include decisions affecting rights, benefits, services, permits, enforcement or employment.
Staff can use approved tools for drafting, research, summaries, translation and analysis without additional approval when the work stays within an approved boundary and a person reviews the output. Uses involving sensitive information or greater authority need further approval.
For all city work, the person responsible must review and verify AI-assisted output before relying on it. A person must still make any decision that requires judgment or discretion.
A public list of tools is required
The policy requires the city to publish and maintain an inventory of approved AI tools and their general purposes. KESQ reported that the city plans to create a public webpage for that list.
As of October 9, the posted materials did not identify the approved tools, define their approved boundaries or say when the inventory would appear. The policy leaves those boundaries to implementation guidelines approved by the city manager.
The city has also decided against a proposed AI-assisted building-plan review system because officials did not believe the technology was ready, KESQ reported.
More authority means more approval
The policy divides AI uses into three risk levels. The level depends on what the AI does and the boundary within which it operates, rather than simply which tool is used. When more than one level applies, the higher one governs.
Moderate-risk uses require the responsible department director’s approval. These include:
- Using sensitive or restricted city information, or inferring sensitive personal data.
- Communicating directly with the public without a person approving each message.
- Establishing or materially expanding a connection that gives AI permission to write to a city system or authority to act.
High-risk uses require both the director’s approval and the city manager’s written approval before they begin. These include:
- Recommending or materially influencing a decision that materially affects a person’s rights, benefits, services, permits, enforcement or employment.
- Communicating or acting without prior human approval in a way that materially affects a person.
- Taking autonomous action that creates a legal or financial commitment for the city, or poses a significant legal, financial, security or operational risk.
City work must use approved AI tools through city-provided accounts and within approved boundaries.
The policy also prohibits using AI to create generalized social scores of people, falsify city records, falsely attribute statements to a person or the city, or bypass security controls or required approvals. Using sensitive city information to train or modify an AI model requires the city manager’s written authorization.
What happens to the records?
The policy distinguishes between AI material used while working and AI material that is itself a city record.
Prompts and outputs used solely as working material are treated as preliminary drafts or notes that the city does not ordinarily retain. Before disposing of them, the person responsible must make sure any information relied on is captured in an appropriate city record kept under the city’s retention schedule.
If a prompt or output is itself the record of a communication or action, it must be retained under that schedule.
Nothing may be disposed of during a legal hold or pending records request, or when preservation is otherwise required—including for pending or reasonably anticipated litigation.
Who the rules cover
The policy covers employees, interns, volunteers, elected and appointed officials, and others authorized to use city-provided AI accounts when their use involves city work or sensitive or restricted city information.
Contractors are governed by their agreements. Personal or campaign use with personal resources and public information falls outside the policy.
The city previously relied on administratively approved guidance written for generative AI tools. Staff said AI had expanded beyond what that guidance covered.
Palm Desert created an Artificial Intelligence Officer position in March 2026. The staff report, prepared by officer Joe Balich, describes the governance policy as that position’s first major deliverable.
Staff said adopting the policy has no direct fiscal impact on the General Fund and does not authorize the purchase of any AI product or service. Future purchases would follow normal budget and procurement processes.
The policy assigns implementation guidelines and schedules or transitional measures for existing AI uses to the city manager. The AI officer is to review the policy and program at least annually.
The vote and the posted record
The October 8 meeting’s regular session began at 4 p.m. in the Council Chamber at 73-510 Fred Waring Drive. The policy appeared as item 13.f on the consent calendar, where routine matters can pass in one motion unless the council removes an item for separate discussion. The resolution says the policy takes effect immediately upon adoption.
KESQ reported the council’s approval that evening. As of the morning of October 9, the city’s agenda page had not been updated with the result. The posted resolution still had blank spaces for its number, adoption date and vote, and no minutes of the October 8 meeting were posted.
The staff report was unchanged. The policy text matched the version reviewed for this story on every provision spot-checked after the meeting.
Update, October 9: This story was updated with KESQ’s report of the vote and the status of the city’s posted record. The city re-posted the attachments after the meeting; the source links below point to the replacement copies.
Update, October 14: Revised the headline and opening, reorganized the article for clarity and added key points. The facts and sources are unchanged. The observations about the city’s posted record remain as of October 9.