Skip to content
Monday, 9 November 2026

AI Ethiquettes

The ethics and etiquette of everyday AI.

Radar

US-CA In force

California AB 2013: Generative AI Training Data Transparency

Developers of public generative AI systems must publish a summary of their training data. In force since 1 January 2026; xAI’s court challenge has so far failed.

Last checked

Status
In force
Jurisdiction
California, US
In force since
Next key date
· Ninth Circuit hears xAI’s appeal
Applies to
Developers of generative AI systems made publicly available in California, released or substantially modified on or after 1 January 2022. Narrow exemptions, for example for security and national-security systems.
Last checked

AB 2013 requires developers of generative AI systems offered to Californians, and released or substantially changed since 1 January 2022, to post documentation about the data used to train them, including a high-level summary of the datasets.

xAI sued to block the law in December 2025. A federal court refused a preliminary injunction on 4 March 2026; the appeal is due to be argued at the Ninth Circuit on 18 November 2026.

Key dates

  1. Signed
  2. Documentation due
  3. Court denies xAI’s request to block the law
  4. Ninth Circuit oral argument

Sources

Changelog

  • Replaced unconfirmed committee-summary link with its canonical address; added Ninth Circuit docket; softened what users can learn
  • Checked; no changes