LACo Supervisors to Consider AI Policy

The Los Angeles County Board of Supervisors is expected to consider a proposed policy on Tuesday. The policy would set rules for the safe, fair and open use of artificial intelligence (AI) across county government.
The policy would set standards, duties and oversight rules for AI systems created by the county, bought from outside companies or used through third-party platforms.
“AI technologies are increasingly used across County departments to improve services, make work more efficient, and help make decisions based on data,” the county’s chief executive officer said in a letter to the board. “As these technologies advance, it is important for the County to have clear rules, protect sensitive information, and make sure AI is used responsibly and ethically.”
The policy would require county departments to follow several key principles: putting people first, keeping systems safe and secure, protecting privacy, being open about how AI is used, treating people fairly, taking responsibility, making sure systems work well, and helping employees do their jobs.
“The AI Policy sets common rules to make sure that any AI technology created, purchased, or used by County departments follows the law, protects County information, maintains public trust, and reduces the risk of unfair bias or misuse,” the letter to the board says.
Under the proposal, AI systems would have to protect personal information, private health records, criminal justice information and other sensitive data from people who are not allowed to access, share or change it.
The policy would also require the county to tell the public why and how it uses its AI systems and to clearly identify work created by AI.
It would ban fully automated decisions that have a major effect on people unless a person meaningfully reviews them. It would also ban secret tracking or social scoring of people, as well as using AI to manipulate people's thoughts or behavior, especially among vulnerable groups.
County workers would not be allowed to upload private or legally protected county information to AI platforms that the county has not approved. They also could not use these platforms to make official decisions or treat AI-generated work as a trusted source without having a person check it independently.
The proposed policy says AI should help workers do their jobs and provide them with training and retraining opportunities. It should not be used to replace county employees or contractors.
Oversight duties would go to the county's chief information officer, chief information security officer, chief privacy officer and information technology managers in each department.
The county says the policy would allow no exceptions.
County Counsel reviewed the proposal, and the Audit Committee approved it on Sept. 16. “There will be no fiscal impact to the County on the proposed AI Policy,” the letter to the board says.
The board delayed discussion of the proposal at its Oct. 6 and Nov. 10 meetings. The county's chief information officer recommends that the board approve it.
