Policies must block confidential information from being input into public generative AI systems; they must also ban unlawful discrimination via AI programs. Court staff and judicial officers must also “take reasonable steps” to confirm the accuracy of …