At its business meeting on February 21, 2025, the Judicial Council got a preview of a new model policy that will help ensure the responsible and safe use of generative AI by California courts. Courts will be able to adopt or modify the model policy as needed.
“This is a positive step in a rapidly developing area,” said Chief Justice Patricia Guerrero, who last year announced the launch of the Artificial Intelligence Task Force to evaluate generative AI for its potential benefits to courts and court users while mitigating risks to safeguard the public. “We must balance the issues you’ve identified: accountability, transparency, confidentiality, and privacy protection.”
The California Judicial Council’s Artificial Intelligence Task Force conducted a survey of courts to understand their current use of AI and their policies regarding generative AI. Here are some key findings from the survey:
– – Generative AI Usage: 19 courts are already using generative AI, and 19 more plan to start using it. Seven courts did not answer this question.
– – Policies: Six courts have a use policy in place, while 21 courts are planning to create one. Many courts are waiting for a model policy from the task force.
– – Model Policy: The task force has developed a model policy for the use of generative AI, which courts can adopt or modify to suit their needs. This policy includes guidelines for reviewing AI-generated material for accuracy, ensuring it is not biased or harmful, and disclosing if AI outputs make up a substantial portion of a work provided to the public.
– – Future Plans: The task force is working on further guidance for courts adopting their own generative AI policies and for judicial officers using AI in their adjudicative roles. They plan to develop a rule of court and a standard of judicial administration on these issues, with an anticipated effective date of September 1, 2025.
“We learned what topics courts intended to cover in their use policies, but also that many were waiting for guidance from the task force and the Judicial Council before drafting their own,” said Justice Mary J. Greenwood, a member of the AI task force. “That helped us establish what the task force should be working on.”
Justice Greenwood shared that the model policy will provide courts with general guidelines for using generative AI in their daily, non-adjudicative duties, which includes direction on:
– – Reviewing generative AI material for accuracy, completeness, errors, and hallucinations
– – Ensuring AI material is not biased, offensive, or harmful
– – Disclosing if generative AI outputs make up a substantial portion of a written or visual work provided to the public
In addition to introducing the model AI policy, the task force also detailed its ongoing work. The task force plans to develop further direction for courts adopting their own generative AI policies, as well as guidance for judicial officers using AI in their adjudicative role.
The task force hopes to develop a rule of court and a standard of judicial administration on these additional issues in the coming months.
You may watch the video recording of this presentation for further details.