Overview
- California will independently review any federal “supply‑chain risk” labels on AI vendors, a shift prompted by the Pentagon’s action against Anthropic that a judge has temporarily blocked.
- The order directs agencies to draft contract standards that stop AI from generating child sexual abuse material or enabling unlawful discrimination, detention or surveillance.
- State workers will receive guidance on watermarking AI‑generated images and videos to help prevent fakes and reduce confusion for the public.
- The Department of General Services and the Department of Technology must propose new vendor certifications within 120 days so bidders can attest to safety practices and civil‑rights protections.
- More than 20 departments are building or using Poppy, a workplace AI assistant, as agencies test tools that help employees and connect people to services under the updated State Digital Strategy.