Home

Tracking California’s Proposed AI Legislation

Alan Kyle / Aug 27, 2024

This tracker, first published July 22, was last updated on Tuesday, August 27.

Illustration of the California State Capitol building in Sacramento.

With a raft of proposed legislation before its legislature, California is in the process of enacting new AI regulations that will affect millions of people both within and outside the state. This bill tracker offers a resource for those interested in understanding the proposed laws, as well as which measures are likely to pass and how they may affect the AI landscape.

The tracker is an Airtable base containing the 30 AI bills that entered the legislative process, and includes tags and categories that can be filtered and grouped. The status of each bill was updated as of August 27.

The deadline for each house to pass bills is August 31. Passed bills will be sent to the governor, who may sign a bill, veto it, or permit it to become law without his signature.

On Thursday, August 15, a significant set of Appropriations Committee hearings resulted in six bills being presumed dead after failing to pass.

They are:

  • AB 1791: Would have prevented the removal of provenance information by social media companies.
  • AB 1814: Sought to regulate facial recognition systems used by law enforcement.
  • AB 1856: Aimed to criminalize the distribution of nonconsensual sexual deepfakes.
  • AB 2877: Proposed to prevent training on personal information of individuals under 16 years old.
  • SB 893: Planned to establish the California Artificial Intelligence Research Hub to facilitate collaboration between government, academia, and industry.
  • SB 933: Intended to add generative AI images to statutes criminalizing Child Sexual Abuse Material (CSAM). (Note: A similar bill, AB 1831, is still in consideration.)

The bills that have passed and await governor action include:

  • AB 2876: This bill, passed last Tuesday, calls for the consideration of media and AI literacy in the school curriculum.
  • AB 2905: This bill, also passed last Tuesday, mandates that calls with pre-recorded AI messages must notify receivers that the messages are AI-generated.

On Monday, three bills unanimously passed a Floor vote and are now on concurrence, meaning they’ve gone back to their house of origin for a final vote now that they’ve been amended:

  • AB 2013: Requires AI developers to share information about training data.
  • SB 896: Requires the California Department of Technology and Office of Emergency Services to produce reports on generative AI usage. Requires notices for certain state uses of generative AI.
  • SB 1288: Convenes a working group to evaluate AI-enabled teaching and learning practices.

17 other bills will also go to concurrence, should they pass their respective second chambers.

One bill, AB 2370, was signed by the governor in early July, and is a measure to prevent community college instructors from being replaced by AI.

SB 1047, known as the Safe and Secure Innovation for Frontier Artificial Intelligence Models Act, is particularly contested (read perspectives on this bill, which would introduce new regulatory requirements and liability on foundation model developers, here and here).

Initial Insights (as of July 22, 2024):

  • Voting Trends: To date, about half of the bills have yet to receive a single “No Vote” in a committee or floor session.
  • Privacy Focus: Privacy is a big theme, with nine bills receiving the "Privacy" tag.
  • Key Lawmakers: Assemblymember Rebecca Bauer-Kahan (D-CA16) leads with four bills introduced, and Senator Josh Becker (D-CA13) with three.
  • Senate vs. Assembly: There are 11 Senate bills and 19 Assembly bills, proportional to the size of each house.
  • The First AI Law: The first law this legislative cycle, AB 2370, was signed by Governor Gavin Newsom, a Democrat, on July 2, 2024.

Authors

Alan Kyle
Alan Kyle is a tech policy professional with experience across AI, trust & safety, and privacy. He specializes in navigating regulatory uncertainty and building systems of information governance. He is a graduate of the UC Berkeley School of Information, where he received a master's degree in Inform...

Topics