Navigating an evolving, layered framework
With Naaia, align your AI systems with California’s expanding regulatory landscape, from training data transparency and content provenance to upcoming legislative obligations.
Assess your complianceA layered, expanding framework
California is regulating AI through a growing set of laws, each covering a specific use of AI rather than one single regulation. This means that requirements build up over time, and companies may have different obligations depending on how they develop or use AI.
Today, the main rules focus on two areas: transparency about training data (explaining what data was used to build generative AI systems) and clear labeling of AI-generated content (such as watermarking and detection tools).
As new laws are introduced regularly, companies need to stay informed and be ready to adapt their AI practices to remain compliant.
AB 2013 – Generative AI training data transparency
AB 2013 requires developers of generative AI systems to publicly disclose information about the data used to train their models. Before each release or significant update, developers must publish a summary covering dataset sources, data types, IP status, and use of personal data — posted on their website and kept current.
The obligation applies to any generative AI system made available to Californians whether free or paid, and covers any person, company, or government agency involved in designing, producing, or substantially modifying such a system.
SB 942 – California AI Transparency Act
SB 942 requires large generative AI providers, those with over one million monthly users publicly accessible in California, to make AI-generated content clearly identifiable.
Providers must offer users a visible label on outputs, embed provenance data into every piece of AI-generated image, video, or audio at the moment of creation, and provide a free public tool that lets anyone verify whether content was produced by their system.
The law also extends to third-party licensees, who must maintain these disclosure capabilities under contract. Non-compliance carries a civil penalty of $5,000 per violation per day.
With Naaia
Map your regulatory exposure in California
Identify which obligations apply based on your AI systems, scale, and distribution model.
Structure training data transparency
Organize and maintain required disclosures on datasets, sources, and data usage.
Operationalize content provenance
Implement labeling, watermarking, and verification mechanisms for AI-generated content.
Stay ahead of regulatory expansion
Track new California AI laws and update your obligations as the framework evolves.
Learn about other regulations & norms
Frequently asked questions
-
What AI transparency and disclosure requirements do California laws impose on businesses?
California has enacted several AI-specific laws that collectively impose significant disclosure obligations.
Key requirements include: publishing documentation of training data (AB 2013), providing tools to detect AI-generated or AI-modified content, disclosing generated content (SB 942), and meeting transparency obligations for automated decision-making systems impacting California residents.
Organizations should conduct a California AI law compliance audit to identify which laws apply to their specific AI use cases, as obligations vary significantly by industry and application type. -
Do California laws require businesses to label or disclose AI-generated content?
Yes. California’s SB 942 (AI Transparency Act), effective January 2026, requires the disclosure of AI-generated content as well as the provision of tools to detect content generated or modified by AI systems. Beyond these specific laws, a broader principle of disclosure applies: consumers must be informed when AI systems make or significantly influence decisions that affect their rights, access to services, or employment. Businesses should review their content generation and decision-making workflows to identify where AI disclosure obligations apply.
-
How should organizations document AI system changes to comply with California AI disclosure laws?
California’s AI disclosure laws require organizations to demonstrate, on request, how their AI systems work and have evolved over time. This means maintaining a change log for each AI system covering: updates to training data, model architecture changes, shifts in intended use cases, and any incidents or anomalies. Documentation should be retained for a minimum period consistent with California consumer rights laws — typically three years for digital records — stored in a format accessible to both technical and legal teams, and designed to support consumer access requests.