Common Themes Emerge in State AI Legislation

Artificial intelligence is the “it” topic in statehouses this year with lawmakers everywhere considering proposals to regulate the rapidly evolving technology.

There’s a lot of legislation to keep track of. Fortunately, the International Association of Privacy Professionals recently took the time to categorize some of the main types of bills aimed at regulating the use of AI in the private sector.

Algorithmic Discrimination Bills

Perhaps the biggest concern legislators have about AI is the potential for it to be discriminatory. As such, there are several bills targeting what’s known as “algorithmic discrimination.”

• California AB 2930 would require AI developers to formally assess and report the impacts of their automatic decision-making programs.

• Connecticut SB 2 would establish reporting requirements for developers who rely on AI as “a controlling factor” in the making of any “consequential decision.”

• Companion measures in Hawaii, HB 1607 and SB 2524, would bluntly outlaw algorithmic discrimination.

• Illinois HB 5116 and HB 5322 would require AI developers to annually assess the impact of any automated decision-making tools they deploy.

• Oklahoma HB 3835 would also require AI developers to assess the impact of their decision-making tools.

• Vermont HB 710 and HB 711 would essentially require developers to use reasonable care to avoid algorithmic discrimination.

Automated Employment Decision-Making Bills

Lawmakers are so concerned about a specific subset of AI-based automated decision-making—within the context of employment—that the IAPP considers it its own category of AI legislation, separate from the more general bills listed above.

• Illinois HB 3773 would prohibit an employer from considering an applicant’s race or zip code when using AI to make hiring decisions.

• Massachusetts HB 1873 would require employees to disclose the use of any automated decision-making system to not only employees but also the state’s Department of Labor & Workforce Development.

• Companion measures in New Jersey, AB 4030 and SB 1588, would prohibit the sale of an AI automated decision-making tool that has not undergone a bias audit.

• New York AB 7859 would require employers and employment agencies to notify job candidates if AI was used to make hiring decisions. Meanwhile, New York SB 5641 would establish criteria for the sale of AI-powered employment-decision tools and SB 7623 would restrict the use of automated employment decision tools as well as electronic monitoring. 

• Vermont HB 114 would also restrict the electronic monitoring of employees and the use of AI employment decision-making.

Bill of Rights and Working Group Bills

Unsurprisingly, another common type of AI legislation is bills that would provide for commissions or working groups to study AI and come up with recommendations for dealing with the new technology.

One example is a pair of companion measures in Hawaii, HB 2176 and SB 2572, which would create the Artificial Intelligence Working Group or the Office of Artificial Intelligence Safety and Regulation within the Department of Commerce and Consumer Affairs to study and regulate AI in the Aloha State.

Then there are a handful of bills that seek to establish an AI Bill of Rights, including the right for residents to know when they’re interacting with AI and how their data is being used. These concepts are often found in the bills already mentioned, but they are also included in legislation like Oklahoma HB 3453 and companion measures in New York, AB 8129 and SB 8209.

Recurring Themes Among State AI Bills

As of late February, a quarter of state legislatures had considered bills in 2024 aimed at regulating the use of artificial intelligence in the private sector. The International Association of Privacy Professionals identified a number of common themes in that legislation, including “algorithmic discrimination, automated employment decision-making, AI Bill of Rights and ‘working group’ bills.”

California Considering Some Unique Approaches

California lawmakers are weighing a trio of bills that don’t fit neatly into any of the categories above.

SB 942 would mandate that certain businesses deploying generative AI systems create an “AI detection tool” that would allow the public to check whether content—text, images, videos, etc.—they provided was made by generative AI. The bill would also require such businesses to include in AI-generated content a disclosure indicating that fact in a form that is “machine detectable” and permanent, or at least difficult to remove. Businesses covered by the bill would also have to register with the state Department of Technology and provide the state with the URLs for their AI detection tools, which the department would add to a “Generative AI Registry.”

“There’s going to be lots of positive uses of this technology, but we’ve seen real risk as well, so my bill fundamentally focuses on transparency,” the bill’s author, Sen. Josh Becker (D), said in a recent interview.

Another outlier from California, AB 3204, would apply to something the measure refers to as “data digesters,” which it defines as businesses that use “personal information to train artificial intelligence.” The bill would require businesses deemed to be data digesters to register with the California Privacy Protection Agency.

One last California bill that doesn’t fit into any of the aforementioned categories is SB 1047. Known as the Safe and Secure Innovation for Frontier Artificial Intelligence Systems Act, the measure would establish “clear, predictable, common-sense safety standards for developers of the largest and most powerful AI systems,” according to a press release from the bill’s author, Sen. Scott Wiener (D).

“Large-scale artificial intelligence has the potential to produce an incredible range of benefits for Californians and our economy—from advances in medicine and climate science to improved wildfire forecasting and clean power development,” Wiener said in the release. “It also gives us an opportunity to apply hard lessons learned over the last decade, as we’ve seen the consequences of allowing the unchecked growth of new technology without evaluating, understanding, or mitigating the risks. SB 1047 does just that, by developing responsible, appropriate guardrails around development of the biggest, most high-impact AI systems to ensure they are used to improve Californians’ lives, without compromising safety or security.”

Another bill defying categorization is Oklahoma HB 3577, which would require health insurance companies to disclose if they use AI in their utilization review process.

Owen Davis, an attorney with Husch Blackwell who helped IAPP categorize the AI bills, said he thinks the evolution of AI legislation will mirror that of the privacy law space, with the federal government not acting and the states “trying to fill the gap.”

Davis recommends that those interested in developments in AI policy carefully watch omnibus legislation pending in California and Connecticut, with the likelihood that momentum will build in this legislative arena.

“It’ll be a trickle,” he said, “and then each year we’ll have more states passing these bills.”

—By SNCJ Correspondent BRIAN JOSEPH

As we’ve previously reported, most states have either introduced or enacted legislation related to AI in the past twelve months. AI continues to be a pressing issue for state lawmakers this year, potentially introducing a host of challenges for businesses. And we don’t foresee that changing any time soon. That is why LexisNexis® State Net® would like to offer you 30 days of AI legislative and regulatory alerts for free.*

Sign up here to Start Receiving Alerts

Disclaimer: LexisNexis® State Net® AI Alert Feed offer is limited to the individual addressee specifically selected for this promotion and is void where prohibited by law or by your employer’s policies. Individual must be a government affairs, legal or compliance professional. Offer expires December 31, 2024. Other restrictions may apply.

—By SNCJ Correspondent BRIAN JOSEPH

Visit our webpage to connect with a LexisNexis® State Net® representative and learn how the State Net legislative and regulatory tracking service can help you identify, track, analyze and report on relevant legislative and regulatory developments.