United States of America: California AI Transparency Act including design requirements enters into force with grace period

On 2 August 2026, the California Artificial Intelligence (AI) Transparency Act enters into force with grace period, introducing design obligations for generative artificial intelligence systems, large online platforms, and capture-device manufacturers. The Act seeks to establish design requirements for generative AI systems, large online platforms, and capture-device manufacturers. The Act imposes technical and design obligations. AI systems must embed provenance data and support latent disclosures in images, video, and audio. It provides that platforms must include user interfaces that disclose provenance information and allow inspection or download. It also provides that capture devices must be designed to embed latent disclosures by default and provide users the option to include additional metadata.