Support Centre

You have out of 5 free articles left for the month

Signup for a trial to access unlimited content.

Start Trial

Continue reading on DataGuidance with:

Free Member

Limited Articles

Create an account to continue accessing select articles, resources, and guidance notes.

Free Trial

Unlimited Access

Start your free trial to access unlimited articles, resources, guidance notes, and workspaces.

California: Bill on Safe and Secure Innovation for Frontier AI Models Act read for second time and amended in Assembly

On June 5, 2024, Senate Bill 1047 on the Safe and Secure Innovation for Frontier Artificial Intelligence Models Act was read for a second time and amended in the California State Assembly. The bill was also re-referred to the Committee on Privacy and Consumer Protection on the same day.

Main provisions

The bill, originally introduced on February 7, 2024, and passed by the Senate on May 21, 2024, aims to authorize a developer of a covered model to determine whether a covered model qualifies for a limited duty exemption before initiating training of that covered model. The bill defines an 'artificial intelligence model' as an engineered or machine-based system that, for explicit and implicit objectives, infers, from the input it receives, how to generate outputs that can influence physical or virtual environments and that may operate with varying levels of autonomy.

Additionally, the bill would require developers, before initiating training of a non-derivative covered model, to comply with various requirements, which include implementing:

  • administrative, technical, and physical cybersecurity protections to prevent unauthorized access to or misuse or unsafe modification of the covered model;

  • the capability to promptly enact a full shutdown of the covered model until the covered model is the subject of a limited duty exemption;

  • all covered guidance; and

  • a written and separate safety and security protocol that:

    • provides reasonable assurance that if a developer complies with its safety and security protocol, either of the following will apply:

      • the developer will not produce a covered model with a hazardous capability or enable the production of a derivative model with a hazardous capability; and

      • the safeguards enumerated in the protocol will be sufficient to prevent unreasonable risk of critical harm from the exercise of a hazardous capability in a covered model;

    • states compliance requirements in an objective manner and with sufficient detail and specificity to allow the developer or a third party to readily ascertain whether the requirements of the safety and security protocol have been followed;

    • describes in detail how the developer will meet requirements; and

    • describes in detail the procedure by which the safety and security protocol may be modified.

Lastly, the bill would create the Frontier Model Division within the Department of Technology which, among other tasks, would review annual certification reports from developers, publicly release summarized findings based on such reports, and assess related fees.

You can read the bill here and track its progress here.

Update: July 3, 2024

Bill re-referred to Assembly Committee on Appropriations 

On July 2, 2024, the bill was re-referred to the Assembly Committee on Appropriations. This follows from its introduction to the State Assembly on May 22, 2024, and passing its second reading in the State Assembly on June 20, 2024. The bill passed the Senate on May 21, 2024. 

You can read the bill here and track its progress here.