As artificial intelligence (AI) becomes more prevalent, governments across the country are grappling with how to regulate and tap into the emerging technology — and so is Nevada.

More than a dozen bills introduced during this year’s legislative session have sought to regulate or expand the use of AI in the Silver State, a contentious issue given the technology's rapid ascent and concerns that it could perpetuate misinformation and lead to adverse outcomes from rent prices to mental health care.

It comes as state agencies are warming up to the idea of using AI — defined as machine-based systems that can create content or make decisions, predictions or recommendations based on existing material — to simplify and expedite certain processes. The Department of Employment, Training and Rehabilitation (DETR) last year began using a Google-run AI program to rule on unemployment appeals (which have long experienced a backlog) with human verification, while the Department of Motor Vehicles is using an AI-driven chatbot to field questions.

In November, the state’s Office of the Chief Information Officer unveiled its policy on the responsible and ethical use of AI in state government. It prohibited any agency-level AI policy from being more lenient than the statewide policy, as well as the use of AI to create discriminatory content or use personal data without anonymization.

The office also announced this week that state agencies would be adopting a Microsoft AI-powered assistant that can be used for tasks including drafting emails and brainstorming ideas, according to a memo obtained by The Indy.

Existing state law also already prohibits using AI to engage in deceptive practices, such as spreading misinformation, or infringing on people’s privacy rights.

But more changes could be coming. Here are the bills moving (or not) through the Legislature related to AI.

The all-encompassing bill

The most expansive AI-related bill is SB199, sponsored by Sen. Dina Neal (D-North Las Vegas), which has been significantly amended.

The bill initially included a prohibition on using AI to develop a police report, but the updated version instead requires police departments to create policies regarding the use of facial recognition software and other AI tools that “may violate civil and constitutional rights of citizens.”

The amended bill also removed a proposed prohibition on using AI to create K-12 lesson plans. Instead, the new version requires that a working group study AI in schools and also develop a policy on the use of AI in lesson planning.

Some aspects of the bill remained in place, including blocking an insurer’s use of patient data to train an AI system without explicit consent of the patient, and prohibiting landlords from using AI-recommended rent prices based on nonpublic data.

In California, AI has been used to increase rent prices by synthesizing nonpublic information to generate recommendations for landlords and their competitors. Some local governments, including in San Francisco, have passed legislation banning the use of AI to set rent prices.

The bill also would significantly increase the state oversight of AI companies, which it defines as someone who develops an AI system and sells it to another person. These arrangements would be required to be reported to the attorney general’s office.

Additionally, the bill would require DETR to survey state employers about how many jobs that have either been lost or are at risk of disappearing because of AI.

Earlier this year, Neal told The Indy that the state’s existing AI policies were “limited.” The bill passed out of its initial committee, but was referred to the Senate Finance Committee.

Helping SNAP recipients?

Nevada could become the next state to use AI to connect recipients of the Supplemental Nutrition Assistance Program (SNAP) to food that would otherwise be wasted.

AB474, sponsored by the Assembly Committee on Ways and Means, would require the state’s Division of Welfare and Supportive Services to use the Smart Shopper mobile app that is being rolled out in Delaware. The bill has received an exemption from legislative deadlines.

The app uses AI to determine how much food is going to waste through retailer data, and then transfers that information to SNAP recipients using food consumption trends that determine who might benefit from the unused food. Recipients could use the app to see discounts on SNAP-eligible food items near them that would otherwise be trashed or go uneaten.

More than eight other states are actively discussing how to implement the app.

“It's a real win-win — food security for the most vulnerable and saving waste,” Bethany Hall-Long, the former lieutenant governor of Delaware who served a brief stint as the state's governor and is working on boosting implementation of the app, said in an interview.

The legislation comes as the Trump administration has cut certain programs that have amounted to a loss of 1.3 million pounds of food being delivered to Nevada food banks.

Proposals on health care, electricity, elections, child porn

The secretary of state’s office is behind AB73, which would require the disclosure of artificial intelligence in campaign materials, which state election officials said is an easy way for bad actors to mislead the public. It passed unanimously out of the Assembly.

Neal has another bill (SB128) that would prohibit health insurers from using or relying on AI to deny a request for prior authorization, or limit the amount of coverage provided for previously approved care. It passed the Senate with support from all Democrats and two Republicans: Sens. Lori Rogich (R-Las Vegas) and John Steinbeck (R-Las Vegas).

A recent survey from the American Medical Association found that 61 percent of physicians were worried that AI was leading to an increase in prior authorization denials.

AB325, sponsored by the Assembly Committee on Government Affairs, prohibits using AI to make final decisions on planning emergency responses and allocating resources, as well as to decide to shut down certain utility services. It unanimously passed on the Assembly floor last week.

Some states have begun using AI to assist with responses to non-emergency calls, and Nevada and other states have also used AI to identify and suppress wildfires, according to a report from the Guinn Center, a Nevada-based policy research organization, that prompted the legislation. However, proponents of the bill said there need to be safeguards in place to ensure that humans have final say over emergency responses.

Senate Majority Leader Nicole Cannizzaro (D-Las Vegas) is sponsoring SB263, which would outlaw computer-generated child pornography, including content created through artificial intelligence. The bill is exempt from legislative deadlines.

An amended bill from Assm. Jovan Jackson (D-North Las Vegas) would prohibit school counselors, psychologists and social workers from using AI to provide care. 

During the hearing on AB406 earlier this month, representatives with the National Association of Social Workers said there is an increase in AI therapy apps with no human involvement, such as Earkick and Sintelly, and there have been instances when people have died by suicide after interacting with AI-powered therapy. 

Jackson said during the hearing that “AI does not provide that emotional intelligence that we need” to provide mental health care.

Bills that are dead (for now)

Several bills that would have imposed certain transparency requirements or limits on AI failed to meet the first committee passage deadline earlier this month, though four of them had similar language as proposals that did pass:

• AB295, a proposal related to AI and prior authorization from Assm. Toby Yurek (R-Las Vegas), has similar language to the other health care-related bill (SB128).
• AB531, brought forward by the Assembly Committee on Education, would have required the state superintendent to create a group to study the use of AI in education. These provisions are also reflected in SB199, the wide-ranging AI bill sponsored by Neal.
• A proposal from Assm. Bert Gurr (R-Elko), AB271, would have prohibited the use of AI in voting equipment and require the disclosure of AI-generated political materials, the latter of which is the subject of the secretary of state office’s bill (AB73).
• AB187, from Assm. Joe Dalia (D-Henderson), mirrors Cannizzaro’s bill to prohibit AI-created child pornography.

Two other AI bills that met their demise were AB537, which would have required the Department of Taxation to disclose the use of AI in tax matters, and SB186, a proposal from Sen. Jeff Stone (R-Henderson) that would have mandated medical entities disclose the use of AI when it is used to generate a written or verbal communication to a patient.