Bill Text: CA AB2013 | 2023-2024 | Regular Session | Amended

NOTE: There are more recent revisions of this legislation. Read Latest Draft
Bill Title: Generative artificial intelligence: training data transparency.

Spectrum: Partisan Bill (Democrat 1-0)

Status: (Passed) 2024-09-28 - Chaptered by Secretary of State - Chapter 817, Statutes of 2024. [AB2013 Detail]

Download: California-2023-AB2013-Amended.html

Amended  IN  Assembly  May 02, 2024
Amended  IN  Assembly  April 22, 2024

CALIFORNIA LEGISLATURE— 2023–2024 REGULAR SESSION

Assembly Bill
No. 2013


Introduced by Assembly Member Irwin

January 31, 2024


An act to add Title 15.2 (commencing with Section 3110) to Part 4 of Division 3 of the Civil Code, relating to artificial intelligence.


LEGISLATIVE COUNSEL'S DIGEST


AB 2013, as amended, Irwin. Artificial intelligence: training data transparency.
Existing law requires the Department of Technology, in coordination with other interagency bodies, to conduct, on or before September 1, 2024, a comprehensive inventory of all high-risk automated decision systems, as defined, that have been proposed for use, development, or procurement by, or are being used, developed, or procured by, state agencies, as defined.
This bill would require, on or before January 1, 2026, and before each time thereafter that an artificial intelligence system or service, as defined, is made available to Californians for use, regardless of whether the terms of that use include compensation, a developer of the system or service to post on the developer’s internet website documentation, as specified, regarding the data used to train the artificial intelligence system or service, as defined. service.
Vote: MAJORITY   Appropriation: NO   Fiscal Committee: NO   Local Program: NO  

The people of the State of California do enact as follows:


SECTION 1.

 Title 15.2 (commencing with Section 3110) is added to Part 4 of Division 3 of the Civil Code, to read:

TITLE 15.2. Artificial Intelligence Training Data Transparency

3110.
 For purposes of this title, the following definitions shall apply:
(a) “Artificial intelligence system or service” or “system or service” intelligence means a an engineered or machine-based system or service that varies in its level of autonomy and that can, for a given set of human-defined explicit or implicit objectives, infer from the input it receives how to generate content and make predictions, recommendations, or decisions influencing a real outputs that can influence physical or virtual environment. environments.
(b) “Developer” means a person, partnership, state or local government agency, or corporation that designs, codes, or produces an artificial intelligence system or service, or substantially modifies an artificial intelligence system or service for use by a third party for free or for a fee.
(c) “Synthetic data generation” means a process in which seed data are used to create artificial data that have some of the statistical characteristics of the seed data.
(d) “Train an artificial intelligence system or service” includes testing, validating, or fine tuning the artificial intelligence system or service.

3111.
 On or before January 1, 2026, and before each time thereafter that an artificial intelligence system or service is made available to Californians for use, regardless of whether the terms of that use include compensation, the developer of the system or service shall post on the developer’s internet website documentation regarding the data used to train the artificial intelligence system or service, including, but not be limited to, all of the following:
(a) A description high-level summary of each dataset the datasets used in the development of the system or service, including, but not limited to:
(1) The source sources or owner owners of the dataset. datasets.
(2) A description of how the dataset furthers datasets further the intended purpose of the system or service.
(3) The number of data points included in the dataset, datasets with estimated figures for dynamic datasets.
(4) A clear definition of each category associated to data points within the dataset, datasets, including the format of data points and sample values.
(5) Whether the dataset includes datasets include any data protected by copyright, trademark, or patent, requiring the purchase or licensure of the data, or whether the dataset is datasets are entirely in the public domain.
(6) Whether the data was datasets were purchased or licensed by the developer.
(7) Whether the dataset includes datasets include personal information, as defined in subdivision (v) of Section 1798.140.
(8) Whether the dataset includes datasets include aggregate consumer information, as defined in subdivision (b) Section 1798.140.
(9) A description of any cleaning, processing, or other modification to the dataset datasets by the developer, including the intended purpose of those efforts in relation to the system or service. If a dataset has datasets have been merged with another dataset, other datasets, the developer shall include the disclosure disclosures required by this section for the original dataset. datasets.
(10) The time period during which the data in the dataset was datasets were collected, including a notice if the data collection is ongoing.
(11) The dates the dataset was datasets were first and last used during the development of the system or service.
(b) A disclosure of whether the system or service used or continuously uses synthetic data generation in its development. A developer may include a description of the functional need or desired purpose of the synthetic data in relation to the intended purpose of the system or service.
(c) A developer shall not be required to post documentation regarding the data used to train an artificial intelligence system or service that has the sole purpose to help ensure security and integrity as defined in subdivision (ac) of Section 1798.140

feedback