SB-313: A Proposal for an Office of Artificial Intelligence in California

On 6 February 2023, Californian State Senator Bill Dodd introduced Senate Bill 313 as part of California’s efforts to regulate the use of AI. While brief, the Bill makes two important contributions, which we outline below.

Establishment of the Office of Artificial Intelligence

Previously, the Department of Technology – supervised by the Director of Technology – was established within the Government Operations Agency. Among the responsibilities of the Director include providing technical direction to agency and department Chief Information Officers to ensure that technology initiatives adopted across California are integrated and that information technology policies and standards are complied with, as well as promoting the alignment and effective management of technology.

SB 313 seeks to extend these efforts, establishing the Office of Artificial Intelligence within the Department to guide the design, use, and deployment of automated systems by a state agency. The aim of this is to ensure that AI systems are deployed and designed in compliance with state and federal regulations, particularly concerning privacy and civil liberties. The Agency will also be responsible for ensuring that bias resulting from AI systems is minimized and that there are equitable outcomes from Californians.

Alignment with the Blueprint for a Bill of Rights

The main risk verticals that SB 313 is concerned with are fairness, transparency, and accountability to ensure that the rights and opportunities of Californians are not adversely impacted by the use of AI within the state.

In particular, SB 313 asserts that AI systems should not discriminate based on race, gender, age, religion, or sexual orientation, and that there should be transparency around the use of AI systems. In addition, they should comply with privacy protection practices and civil liberties.

Such priorities are aligned with those of the White House’s Blueprint for an AI Bill of Rights, which was published in October of 2022. Indeed, the Blueprint focuses on the prevention of discrimination, data privacy, and giving notice and explanation (resulting in greater transparency), in addition to ensuring that there are human alternative and safeguards against unpredicted harms, with an overarching goal of making AI systems safe and effective. However, both the Blueprint and SB 313 currently lack specific actions and enforcement guidelines in their current form.

Future iterations

Having just been introduced, the text is very brief and specific details are missing. However, Dodd plans to amend the text to require that Californian’s are informed about their interactions with AI-relevant communications, and the Bill is likely to continue to evolve further.

To find out how Holistic AI can help you with compliance with AI regulations, get in touch at we@holisticai.com