Across the United States, policymakers are demonstrating a clear impetus to regulate AI, with laws being proposed at the local, state, and federal levels.
Among the states leading this race is California, where multiple initiatives have been proposed. Although California does not yet have a dedicated Artificial Intelligence Act, it has proposed both horizontal and sector-specific laws targeting HR Tech, as well as a focus on state agency use of AI.
In this blog post, we give a high-level overview of California’s proposed AI legislation and discuss the potential for a California AI Act.
California AB302 was introduced on 26 January 2023 and passed on 14 September 2023 before being enrolled on 18 September 2023. The bill adds Section 11546.45.5 to the Government Code to establish the Department of Technology within the Government Operations Agency, where the Director of Technology within the newly formed department will report directly to California’s Governor on issues relating to automated decision systems (ADTs). These systems are defined as:
“a computational process derived from machine learning, statistical modelling, data analytics, or artificial intelligence that issues simplified output, including a score, classification, or recommendation, that is used to assist or replace human discretionary decision making and materially impacts natural persons”
High-risk automated decision systems are those that are used to make decisions that materially impact access to, or approval for, housing or accommodations, education, employment, credit, health care, and criminal justice. The law, therefore, does not exclusively regulate AI and instead has a wider scope that applies to various decision-making technologies used in high-stakes contexts.
The law requires the Department of Technology to conduct a comprehensive inventory of all high-risk automated decision systems that have been proposed for use, development, or procurement by, or are being used, developed, or procured by, any state agency in California. This inventory should include:
This inventory must be compiled on or before 1 September 2024, giving the Department of Technology around one year to catalogue systems being used by California State agencies. Additionally, annually starting 1 January 2025 at the latest, the Department of Technology must submit a report of the inventory to the Assembly Committee on Privacy and Consumer Protection and the Senate Committee on Governmental Organization. This requirement, however, will be discontinued from 1 January 2029.
In addition to AB 302, California lawmakers have also proposed SB-313 to regulate the use of AI by state agencies. Introduced on 6 February 2023, the has been held in committee and under submission after being referred to the Committee on Appropriations so may not make it any further in the legislative process.
Nevertheless, the law seeks to enact the California AI-ware Act, which would establish the Office of Artificial Intelligence within the Department of Technology.
In particular, the Office would be given the power and authority to guide the design, use, and development of automated systems by California State Agencies to ensure that AI systems are designed and deployed in a way that is compliant with state and federal laws and regulations to minimise bias and promote equitable outcomes.
The Act would also impose transparency and notification requirements, where any state agencies using generative AI to disclose to individuals, either online or telephonically, declare the use of the generative AI clearly and conspicuously. State agencies using generative AI would also be required to provide (a link to a page with) clear instructions on how they can directly communicate with someone from the state agency.
Outside of state agency use of AI, AB 331 was introduced in January 2023 by California assembly member Rebecca Bauer-Kahan to prohibit algorithmic discrimination through automated decision tools (ADTs). These are defined as:
“a system or service that uses artificial intelligence and has been specifically developed and marketed to, or specifically modified to, make, or be a controlling factor in making, consequential decisions”
In particular, AB 331 seeks to regulate ADTs used for making consequential decisions, or those that have a legal, material, or significant effect in terms of access to, cost of, terms of employment, worker management, or self-employment; education and vocational training, housing or lodging, essential utilities, family planning, health care or health insurance, financial services, the criminal justice system, legal services, voting, and access to benefits or services or the assignment of penalties.
Importantly, the bill would impose obligations for both developers of ADTs and deployers of ADTs, where both would be required to conduct an impact assessment and at least one of them will be required to establish, implement, and maintain a governance program to map, measure, manage, and govern the reasonably foreseeable risks. Additionally, they must make a clear AI policy publicly available that summarises the types of ADTs in use or made available and how the risks associated with each system are managed.
Furthermore, deployers of ADTs must notify any natural person affected by the decision that an ADT is being used including a statement of the purpose of the ADT, contact information, and a plain language description of the ADT. To assist this, developers must provide deployers with a statement of the intended use of the ADT and documentation regarding the known limitations of the tool.
More recently, SCR17 was passed on 15 August 2023 before being enacted on 23 August 2023 to affirm the California Legislature’s commitment to the White House’s Blueprint for an AI Bill of Rights, which is based on five key principles:
With SCR 17, the Senate has expressed a commitment to examining and implementing the principles in its legislation and policies around automated systems in California.
Assembly Bill 331 is the closest that California has got to an AI Act and, like the EU AI Act, it takes a horizontal approach to legislation that targets multiple use cases. However, AB331 has been held under submission after being referred to the Committee on Appropriations, so may not progress out of Committee.
Nevertheless, the AI policy landscape is evolving rapidly and it is clear that Californian policymakers are increasingly recognising the need for AI legislation and are targeting multiple critical applications.
Proposals around the globe, such as the EU AI Act, will, however, have worldwide implications, including for developers of AI tools based in California.
With the wave of upcoming regulations to navigate, it is vital to take steps early. Schedule a call with our experts to find out how you can manage risks and adopt AI with confidence.
DISCLAIMER: This blog article is for informational purposes only. This blog article is not intended to, and does not, provide legal advice or a legal opinion. It is not a do-it-yourself guide to resolving legal issues or handling litigation. This blog article is not a substitute for experienced legal counsel and does not provide legal advice regarding any situation or employer.
Schedule a call with one of our experts