Law / United States / California

Generative AI Training Data Transparency Act (AB 2013)

Cal. Civ. Code Sections 3110 and 3111

In force since , under challenge in court.

An AI training data rule binding public and private bodies.

Criminal exposure
No
Private right of action
No
How the hook was established
express
What makes it apply
Market targeting
Obligation class
Disclosure
Audit expectation
none
Court action

A suit challenging this law is pending. No court has blocked it. Premium data

What it requires

  • Post training-data documentation on your website before making a generative AI system publicly available to Californians, and again before any substantial modification
  • Cover the datasets and their sources and owners, an approximate count and description of the data points, whether the data includes copyrighted or personal information, and whether synthetic data generation was used
  • The duty reaches systems first made publicly available on or after

These lines are LexLint's reading of the law's text.

In the frameworks

NIST AI RMF
  • MAP 2.3Scientific integrity and TEVV considerations are identified and documented, including those related to experimental design, data collection and selection (e.g., availability, representativeness, suitability), system trustworthiness, and construct validation.
NIST Privacy Framework
  • ID.IM-P6Data elements within the data actions are inventoried.

Our reading of where this law's duties sit in each framework. It never says that running a control meets the law. How we read a framework against the law

Who enforces it

Enforcement body

No dedicated regulator, civil penalty, or private right of action is stated in Title 15.2 (Sections 3110 and 3111) itself; the full text of Sections 3110 and 3111 contains no enforcement, remedy, or penalty provision. A violation could be reached only through the general Unfair Competition Law (Bus. and Prof. Code section 17200 et seq.), enforceable by the Attorney General or a district attorney.

Settledness

A Ninth Circuit appeal, No. 26-1591, from the Central District of California, No. 2:25-cv-12295, was pending with an amicus brief filed on .

As of
Under challenge
Yes
Open questions
  • Does a "high-level summary of the datasets" require a developer to name each dataset it used, or only to describe categories of data?
  • Does a system "made publicly available to Californians" include one offered only to business customers through an application programming interface?

What it makes you log

Who may demand the log

Public

What the log must hold

Data lineage

Logging duty

Section 3111 requires a developer to post training-data documentation on its own website before the system is made available and before each substantial modification. It never uses the words log, record, event recording, or audit trail; it requires posting documentation, which is the implicit category's own verb, and the document must persist on the website rather than being a one-time disclosure at the point of output.

Kind
Implicit
As of
Provision
Section 3111
Trigger
content_generation

What this law does

Drafted with AI

Drafted with AI from the cited sources under the direction of UnGovr staff. UnGovr holds editorial responsibility for this page. How this site is made

Research summary

Legal information, not legal advice. This is LexLint's own research summary of a public legal source, and it creates no attorney-client relationship. For decisions that matter, consult qualified counsel in the relevant jurisdiction. About LexLint

A developer of a generative AI system made publicly available to Californians must post on its website, before making the system available and before any substantial modification, documentation describing the training data: a high-level summary of the datasets, their sources and owners, an approximate count and description of data points, whether the data includes copyrighted or personal information, and whether synthetic data generation was used.

The posting duty reaches systems first made publicly available on or after , though it did not attach until the operative date, now past. This corrects the derived citation on record, Civil Code Section 22756 et seq., which does not exist in Civil Code; confirmed against leginfo, the correct codification is Title 15.2 (commencing with Section 3110), Sections 3110 and 3111.

When LexLint raises it

When your app profile says your app trains models or generates content with AI.

Back to the example  ·  Lint your app

Page updated .