Public bodies would set rules for employees using artificial intelligence

Public bodies would have to set rules for employees who use artificial intelligence, automated decision tools, and technology resources. This covers state agencies and listed local public bodies, including counties, municipalities, school districts, charter schools, and public colleges. The proposed effective date is July 1, 2026.

Each public body's rules would have to address data security, permitted tool uses, human review, and safeguards against bypassing security systems. These rules would apply to public bodies' use of the tools and resources by employees. An authorized employee would have to make the final decision when a consequential decision affects a person's access to or the cost or terms of education, jobs, loans, housing, health care, insurance, or legal services. Bypassing security or system safeguards would be prohibited, except during an authorized security test or with permission from the designated security officer. The proposed effective date is July 1, 2026.

Public bodies would train employees on cybersecurity and tool-use policies and on appropriate use of artificial intelligence and automated decision tools in decisions of any kind. This applies to employees of every public body covered by the bill. The proposed effective date is July 1, 2026.

State agencies and other listed state bodies would provide their policies to the public upon request. This applies to executive, judicial, and legislative state agencies and listed state boards, commissions, departments, and instrumentalities. The proposed effective date is July 1, 2026.