Legislating the Disclosure of AI Use

Legislative policies requiring the disclosure of the use of AI are developing side-by-side along the emerging standards in AI provenance—although they’re not perfectly in sync.

The industry’s C2PA standard has been evolving for a number of years, while the nation’s first significant state law requiring the disclosure of AI, the California AI Transparency Act (SB 942), was just passed and signed into law in Sept. 2024.

While provenance products like Content Credentials are already being used by a handful of companies, they represent the exception not the rule. Most AI-generated images are being disseminated with no way of knowing their authenticity.

Why ai disclosure laws are needed

State AI disclosure laws like SB 942 incentivize the creation of ethical AI tools and products while driving unethical and malicious actors out of the marketplace. They establish appropriate guardrails that encourage high-quality innovation and discourage the spread of misinformation and deceptive imagery.

Here’s how they work.

The California AI Transparency Act (SB 942) takes full effect on Jan. 1, 2026. As of that date, large AI providers serving consumers in California must make available an AI detection tool on the provider’s website. 

That AI detection tool will function as a kind of decoding device, reading the metadata embedded in an AI object’s code and disclosing whether the digital content was created or modified with AI. In the language of the Act, this embedded metadata acts as a “latent disclosure” attached to the digital content but not apparent to the naked eye. 

That requirement is actually less disclosure than what a tool like Content Credentials offers its users—but it establishes a fair baseline and a reasonable deadline for all AI developers.

state AI Disclosure bills coming in 2025

The Transparency Coalition’s leaders and subject matter experts are currently consulting with legislators in a number of states as they prepare AI disclosure bills to be introduced in early 2025. Many state leaders have expressed interest in proposals similar to SB 942, which would both protect their state’s citizens and reinforce an emerging national standard for AI developers.

If you are a legislator, staff member, policy maker or thought leader and would like to get in touch with TCAI, please reach out via the form below.

Previous
Previous

How to Format and Require Data Disclosures

Next
Next

Data Privacy in the Age of AI