State AI Law · California · 3 statutes on the books
AI Laws in California
Every AI-related law currently on the books or recently enacted in California. Each entry below shows the effective date, regulated scope, key triggers, enforcement authority, and penalty exposure — plus a link to the statute text.
Laws at a glance
California's AI statutes cover: AI detection, AI watermark, LLM, candidate, content authenticity, dataset disclosure, election deepfake, foundation model, generative AI, large online platform, model developer, political advertisement. Details for each law follow below. Note that individual statutes frequently have phased effective dates — pay close attention to which provisions apply when for your entity.
Each law, in detail
Effective 2026-01-01
CA AB 2013 (Generative AI Training Data Transparency)
Scope: Requires developers of generative AI models or services made available to Californians to publicly post documentation describing training datasets, including sources, ownership, personal information, and copyrighted data.
Key triggers / concepts
generative AItraining datafoundation modelLLMdataset disclosuremodel developer
Enforcement authority
California Attorney General
Penalties
Injunctive relief and civil penalties under California UCL
Read the statute on the state site →
Effective 2026-01-01
CA SB 942 (AI Transparency Act)
Scope: Requires covered providers of generative AI systems with over 1 million California users to offer a free AI detection tool and embed both manifest and latent disclosures in AI-generated content.
Key triggers / concepts
generative AIsynthetic contentAI watermarkAI detectionprovenancecontent authenticity
Enforcement authority
California Attorney General
Penalties
Up to $5,000 per violation per day
Read the statute on the state site →
Effective 2026-01-01
CA AB 2655 (Defending Democracy from Deepfake Deception Act)
Scope: Requires large online platforms to block or label materially deceptive AI-generated content about elected officials and candidates during specified election periods.
Key triggers / concepts
election deepfakepolitical advertisementsynthetic medialarge online platformcandidate
Enforcement authority
Candidates, elected officials, and AG (private right for injunctive relief)
Penalties
Injunctive relief and damages; recently enjoined in part by federal court (Kohls v. Bonta), pending appeal
Read the statute on the state site →
If your AI system is high-risk, the EU AI Act may also apply
If you operate in the EU or process EU residents' data, the EU AI Act classifies systems as prohibited, high-risk, or limited-risk. Our free 8-question classifier returns the tier + Articles that apply.
Check EU AI Act applicability →
Get AI Governance Daily in your inbox
Every weekday morning: new state/federal AI legislation, enforcement actions, and regulatory guidance. 5-min read, free.