AI Bills by the Numbers – The 2024 California Legislative Session

A lot happened in the California Legislature this year, especially on the hot-button issue of AI. Throughout 2024, our Advocacy team tracked more than 60 AI bills as they made their way through the legislative process. Read our blog AI, the California Legislature, & the will of the Governor for the nitty gritty on all of them. 

With so many bills, it’s hard to understand where AI regulation is going in California. So here are seven charts to help capture how AI bills, their authors, and vote results took shape this legislative session.

AI bills breakdown

In our previous blog, AI, the California Legislature, & the will of the Governor, we looked at all the AI bills we tracked, whether they were signed, vetoed, or failed to make it to the Governor’s desk altogether. It’s clear that Disinformation bills that addressed issues like deepfakes and election interference were favorable for the Governor; he signed eight bills on the topic. Bills on AI & Education, as well as Data, Surveillance, and Privacy, were signed into law as well.

AI was such a popular issue during this session that it felt like almost every legislator in the Capitol was carrying a proposal on the topic. Across both houses, 32 legislators carried at least one AI bill. The above chart shows which legislators carried two or more bills on AI. Senator Becker (D-Menlo Park), Senator Wahab (D-Hayward), and Senator Berman (D-Palo Alto) had 3 or more bills signed. Becker’s bills were on Healthcare (SB 1120), Education (SB 1288), and Disinformation (SB 942). Berman’s bills were on Kids (AB 1831), Education (AB 2876), and Disinformation (AB 2655). Wahab’s bills were on Kids (SB 933 [failed] & SB 1381 [signed]) and Disinformation (SB 981 & SB 926). Nineteen legislators carried a single bill on AI; those authors are not displayed on the chart, but they were: Aguiar-Curry, Ashby, Bradford, Calderon, Carrillo, Cortese, Dodd, Gonzalez, Irwin, Limon, Muratsuchi, Nguyen, Pacheco, Pellerin, Skinner, Smallwood-Cuevas, Ta, Valencia, Weber, Wiener.

How the bills played out in committees

In the California Legislative process, after a bill is introduced by a legislator, it is sent to a policy committee based on its issue area for deliberation, amendment, and approval (or denial) by the committee.

In the Assembly Privacy Committee, chaired by Asm. Bauer-Kahan (D-San Ramon), Asm. Dixon (R-Newport Beach), Asm. Hoover (R-Folsom), and Asm. Patterson (R-Rocklin) voted no or withheld their vote (known as “no vote recorded” or NVR in the legislature) on the most proposals compared to their Democratic peers on the committee. Asm. Lowenthal (D-Long Beach), Asm. Ortega (D–Hayward), and Asm. Ward (D-San Diego) consistently voted yes, approving bills in the committee 100% of the time.

The above chart provides the analysis for the Senate Judiciary Committee, a committee that most AI bills were referred to given that the Senate does not have a Privacy Committee like the Assembly does. Similar to the Assembly Privacy Committee, the Republican members of the Senate Judiciary Committee voted no or withheld their vote most frequently on the AI bills they saw, including Sen. Niello (R-Roseville) and Sen. Wilk (R-Lancaster). Sen. Allen (D-El Segundo), Sen. Ashby (D-Sacramento), Sen. Durazo (D-Los Angeles), Sen. Roth (D-Riverside), Sen. Umberg (D- Santa Ana) who is the Chair of the Committee, Sen. Wahab (D-Hayward), and Sen Wiener (D-San Francisco) all voted yes on every AI bills that went before them.

What passed by a landslide—and what didn’t

Overall, the floor vote margins for most AI bills were wide, with plenty of votes to spare to get over the win line. The chart above displays how many spare votes a bill received on the Senate floor beyond the 21 needed to pass. Bills with the tightest margin for victory are to the left of the chart, and the bills with a nearly unanimous vote are to the right of the chart. 

On the Senate floor, the bill with the tightest margin of victory were:

• AB 1777 (Ting) – Safety Requirements for Autonomous Vehicles, with just two votes beyond the 21 needed.
• SB 1446 (Smallwood Cuevas) – Retail Theft, Automated Checkouts, and Tech Rollouts at Work, with four votes beyond the 21 needed, and subsequently did not move beyond the Asm. Appropriations Committee.
• SB 915 (Cortese) – Local Approval for Autonomous Vehicles, with five votes beyond the 21 needed; it did not get voted on in the Assembly floor.

Twelve bills had between 8-15 spare votes for victory, 20 bills had 16 -19 spare votes, and six bills had a Senate vote but died before hitting the Assembly floor. 

Similar to the Senate, most bills enjoyed a comfortable margin in the Assembly. The chart above displays how many spare votes a bill received beyond the 41 needed to pass. 

On the Assembly floor, the bills with the tightest margin of victory were:

• SB 976 (Skinner) – Protecting Kids from Social Media Addiction, with six spare votes beyond the 41 needed.
• SB 1047 (Wiener) – Safe and Secure Innovation for Frontier Models, with seven spare votes beyond the 41 needed.
• AB 1791 (Weber) – Digital Content Provenance, with nine spare votes beyond the 41 needed.
• AB 2930 (Bauer-Kahan) – ADT Bias Impact Assessments, also with nine spare votes beyond the 41 needed.

Sixteen bills had between 12-29 spare votes for victory, seventeen bills had 30-36 spare votes for victory, and six bills had an Assembly vote but died before hitting the Senate floor.

Vetoes

The Governor signed into law 24 of the AI bills we tracked during the session, but he vetoed a few as well. When the Governor vetoes a bill, he usually issues a Veto Message describing the reasons for rejecting the proposal. Often, he rejected proposals he felt conflicted with his Executive Order, or that cost the state too much in a tough budget year. Above are just a few excerpts from his veto messages; read on to learn what he vetoed. 

SB 1220 (Limon) prohibits state and local agencies from using public benefit-related call center services that use artificial intelligence (Al) or automated decision-making systems (ADS) that eliminate or automate the core job function of a worker. This bill also extends to local governments an existing state requirement that public benefit-related call center services be performed solely by workers employed in California. You can read Newsom’s full veto message here. 

SB 1047 (Wiener) would have required developers of large artificial intelligence (Al) models and those providing the computing power to train such models to put certain safeguards and policies in place to prevent catastrophic harm. The bill would also establish the Board of Frontier Models—a state entity—to oversee the development of these models. Read the veto message here. 

AB 3048 (Lowenthal) would have required internet browsers and mobile operating systems to include a setting that California consumers can use to signal to businesses with which they interact that they wish to, first, opt out of the sale or sharing of their personal information and second, limit use of their sensitive personal information. Newsom’s veto message is here. 

SB 892 (Padilla) would have tasked the California Department of Technology (CDT) to create an automated decision system procurement standard that complies with various requirements specified in the bill. Beginning on January 1, 2027, every state contract for an automated decision system would have had to comply with these new regulations. The bill was one TechEquity supported; you can read the veto message here. 

AB 1949 (Wicks) would have amended the California Consumer Privacy Act (CCPA) to prohibit the sale, sharing, disclosure, or use of minors’ personal information unless the minor’s parent or guardian (for those under 13) or the minor themselves (for those aged 13-18) consents. Read the veto here.

AI legislation in 2025

The flight over AI isn’t over in the Legislature, and there’s sure to be another crush of AI bills that will move their way through the process in 2025. If you want to be part of the movement for people-first AI, and make sure you stay up to date on the latest AI news, policy, and regulatory developments, join us!