AI Oversight Measures Advance at State Capitol

Published On: June 20, 2024

Several AI-related measures pending at the State Capitol have advanced, moving closer to the governor’s desk for approval, including Senate Bill 1047 by Senator Scott Wiener (D-San Francisco) which proposes to create the Safe and Secure Innovation for Frontier Artificial Intelligence Models Act.

With amendments to the bill, the Assembly Privacy & Consumer Protection Committee passed SB 1047 on Tuesday which would establish safety standards for the development of large-scale artificial intelligence systems. The bill, which received unanimous support (8-0), now moves to the Judiciary Committee.

“By taking a light touch approach to regulation that focuses exclusively on the largest companies building the most capable models, SB 1047 puts sensible guardrails in place against risk while leaving startups free to innovate without any new burdens. We know this bill is a work in progress, and we’re actively meeting with stakeholders and seeking constructive input. We accept the amendments introduced by the Committee today and look forward to continued conversations to improve the bill as it advances,” said Senator Wiener said in a press release.

According to the press release from Senator Wiener, the committee introduced amendments to the bill, which include:

  • Giving companies more flexibility in how they meet their responsible development obligations while still holding them liable if irresponsible behavior leads to catastrophic harm.
  • Allowing a state agency to change the compute threshold for a model to be covered under the bill starting in 2027, rather than keeping it fixed in statute at 10^26 flops. The requirement that the model cost at least $100 million to train cannot be changed.
  • Requiring companies get third-party safety audits by 2028.
  • Strengthen whistleblower protections.

Other AI-related measures include:

AB 2013 (Irwin D) – Artificial Intelligence: Training Data Transparency
  • Summary: Extends existing requirements for inventorying high-risk automated decision systems to include AI system developers. Developers must post documentation about the datasets used to train AI systems online by January 1, 2026.
  • Status: Amended and re-referred to the Senate Committee on Judiciary as of June 17, 2024.

AB 2877 (Bauer-Kahan D) – California Consumer Privacy Act of 2018: Artificial Intelligence Training

  • Summary: Prohibits the use of personal information from consumers under 16 years old for AI training without explicit consent. Requires deidentification and aggregation of such data if consent is given.
  • Status: Amended and re-referred to the Senate Committee on Judiciary as of June 17, 2024.

AB 2930 (Bauer-Kahan D) – Automated Decision Tools

  • Summary: Mandates impact assessments for automated decision tools and requires compliance with civil rights regulations. Establishes penalties for discrimination resulting from these tools.
  • Status: Amended and re-referred to the Senate Committee on Judiciary as of June 3, 2024.

AB 3211 (Wicks D) – California Provenance Authenticity and Watermarking Standards

  • Summary: Introduces watermarking requirements for AI-generated content and mandates transparency about the provenance of synthetic content on large online platforms starting from 2025.
  • Status: Passed and amended in the Senate Committee on Governmental Organization as of June 18, 2024.

SB 892 (Padilla D) – Public Contracts: Automated Decision Systems: AI Risk Management Standards

  • Summary: Requires the Department of Technology to establish risk management standards for AI in public contracts, including thorough risk assessments and procedures for adverse incident monitoring.
  • Status: Referred to the Assembly Committee on Privacy and Consumer Protection as of June 3, 2024.

SB 893 (Padilla D) – California Artificial Intelligence Research Hub

  • Summary: Proposes the establishment of the California Artificial Intelligence Research Hub to promote collaboration between government, academia, and private sector for AI research and development.
  • Status: Referred to the Assembly Committee on Privacy and Consumer Protection as of June 3, 2024.

SB 896 (Dodd D) – Generative Artificial Intelligence Accountability Act

  • Summary: Requires government agencies to report on the benefits and risks of generative AI and mandates risk assessments for its use in critical energy infrastructure.
  • Status: Amended and re-referred to the Assembly Committee on Privacy and Consumer Protection as of June 4, 2024.

SB 942 (Becker D) – California AI Transparency Act

  • Summary: Requires AI content providers to disclose AI-generated content and implements fines for non-compliance. The Attorney General is responsible for enforcement.
  • Status: Passed and amended in the Assembly Committee on Judiciary as of June 18, 2024.

 

 

About the Author: Staff

Contact us, share tips and news: info@techca.org. Contributors to this site include writers, analysts and researchers who occasionally use AI tools to perform routine tasks, such as analyzing and transcribing documents, and checking grammar and spelling.