Generative AI Training Data Transparency Law (California AB 2013) state silhouette

Generative AI Training Data Transparency Law (California AB 2013)

Effective Date: 2026-01-01

California AB 2013 requires developers of generative AI systems made available to Californians to post public documentation about the data used to train those systems. It is a transparency law rather than a restriction on what data can be used, but the disclosure it demands is detailed.

What California AB 2013 Requires

The published documentation must summarize the training datasets, including their sources, whether they contain copyrighted or personal information, whether the data was purchased or licensed, and the time period over which it was collected.

Who Is Covered and When

The law applies to generative AI systems released or substantially modified on or after January 1, 2022, and the disclosure obligation began January 1, 2026. Coverage is not limited to the largest labs. Any company shipping a generative AI product into California should confirm that its training data documentation is published and covers the required categories. For smaller vendors that fine-tune or substantially modify existing models, that modification can itself trigger the obligation.

Source: California Legislative Information: AB 2013

Report a violation: If you believe a business is violating this law, you can file a complaint with the California Attorney General Consumer Protection.

More Rules & Regulations