As more and more companies launch “AI” businesses, California’s legislature is working to keep up with laws to protect the public. While there are many new laws out there regarding use of deep fakes, celebrity images, political advertisements and more, this article addresses California’s recent AI Transparency Act, which went into effect on August 2, 2026.
California’s AI Transparency Act, codified at California Business and Professions Code sections 22757 et seq., became operative on August 2, 2026, after the Legislature amended the original effective date through AB 853. The law is one of the nation’s first comprehensive efforts to require transparency and provenance disclosures for AI-generated content.
Although most businesses will not be directly regulated by the Act (since on its face it impacts only organizations with 1M or more monthly users) it contains provisions that may extend obligations to many other companies that use GenAI platforms, including many of our clients. In particular, organizations that use third-party generative AI platforms to create marketing materials, videos, images, audio content, or other communications should understand how the law may affect their operations.
Who Is Directly Covered by the Act?
The Act primarily applies to “covered providers,” defined as entities that create, code, or otherwise produce a publicly accessible generative AI system with more than 1M monthly visitors or users within California.
In practical terms, the law is aimed primarily at large AI platform providers rather than businesses that merely use those tools. Companies that develop and operate large-scale generative AI systems may therefore bear the primary compliance obligations under the statute.
Required AI Content Disclosures
The Act requires covered providers to implement both visible and embedded disclosures for AI-generated or AI-altered image, video, and audio content.
Manifest (Visible) Disclosures
Covered providers must offer users the option to include a clear and conspicuous disclosure indicating that content was generated or altered using AI. The disclosure must be understandable to a reasonable person and, where technically feasible, difficult to remove.
Latent (Embedded) Disclosures
Covered providers must also embed provenance information within AI-generated content. To the extent technically feasible, this information may include:
- The name of the AI provider;
- The name and version of the AI system;
- The date and time of content creation or alteration; and
- A unique identifier associated with the content.
These disclosures must be detectable by the provider’s AI detection tools and aligned with accepted industry standards.
Why Businesses Such as Yours That Use AI Should Still Pay Attention
Even if a company is not itself a covered provider, the Act contains provisions that extend downstream through licensing arrangements.
Specifically, section 22757.3(c) requires covered providers to contractually require third-party licensees to maintain the AI system’s capability to include required disclosures in generated or altered content. Under the statute, if a provider determines that a licensee has modified a licensed AI system so that it can no longer generate the required disclosures, the provider must revoke the license within 96 hours of discovering the issue, and the licensee must cease using the system.
As a result, businesses that use commercially licensed AI tools should expect increasing contractual obligations relating to AI transparency and disclosure functionality.
Practical Compliance Considerations
Organizations that use generative AI should consider taking several practical steps:
- Review vendor agreements to watch for provisions requiring customers to preserve disclosure, attribution, provenance, watermarking, or other transparency features embedded by the provider.
- Preserve disclosure functionality and avoid modifying AI workflows in ways that could disable, remove, strip, or circumvent disclosure mechanisms built into licensed AI systems.
- Evaluate content workflows by having marketing, creative, and communications teams inventory their AI-assisted content creation processes and ensure that any embedded provenance data or disclosures are not inadvertently removed during editing, conversion, compression, or publication.
- Monitor evolving requirements since the AI regulatory landscape remains highly dynamic. Organizations should continue monitoring developments at both the California and federal levels, as well as emerging industry standards relating to content provenance, watermarking, and AI-generated media.
Looking Ahead: Potential Changes to the Act
Businesses should also be aware that California lawmakers continue to consider additional AI legislation that could expand transparency obligations beyond the current framework. For example, proposed legislation such as SB 1000 would potentially broaden the scope of regulated entities by eliminating the 1M user threshold and could eliminate or modify existing applicability thresholds. Because legislation often undergoes significant amendment during the legislative process, the ultimate requirements remain uncertain until a bill is enacted. Accordingly, organizations should monitor legislative developments and be prepared for additional compliance obligations in the future.
Key Takeaway
While California’s AI Transparency Act is directed primarily at large generative AI providers, organizations that use AI-powered tools are not entirely outside its reach. Through contractual requirements imposed on licensees, the Act may effectively require businesses to preserve disclosure and provenance features embedded in AI-generated content. Companies that rely on generative AI for marketing, design, communications, or media production should review their AI governance practices now to ensure they are not inadvertently creating compliance risks as California’s transparency framework continues to evolve.
Scherer Smith & Kenny LLP remains available to assist you with these and any other business, intellectual property and employment law-related questions you may have. For additional information, please contact Brandon Smith at brandon@sfcounsel.com
– Written by Brandon D. Smith