Generative artificial intelligence — the kind that generates text or images, based on prompts — learns a lot like humans do: It picks up language, with its vocabulary growing with every human interaction, and it stores those encounters as data so it can respond more appropriately in the future.
Members of the Utah Legislature want the humans in Utah’s business sector and the state’s department of Commerce to learn about AI the same way.
The Senate Business and Labor Committee, on a unanimous vote, sent SB149 ahead to the full Senate on Thursday. The bill, if it becomes law, would establish an AI “learning lab” for businesses and state regulators to keep tabs on the newest technology and trends in artificial intelligence — so the state can better understand how to regulate AI.
The bill would impose some immediate rules on consumer-facing AI — because while AI can already act human, that doesn’t make it human.
“We want to put some guardrails in place now,” Margaret Busse, executive director of the Utah Department of Commerce, told the committee.
Under SB149, businesses that use generative AI could be held liable if that AI deceives consumers in violation of Utah’s consumer protection laws. For example, if a chatbot misleads a consumer into buying a product, the chatbot is not responsible for that lie — the company behind it is.
“You can’t use AI as a defense,” Busse said.
The bill would also require any consumer-facing generative AI, such as chatbots or text messages, to answer honestly if asked, “Are you human?” AI in certain licensed industries — including health care, mental health and finance — would be required to disclose that it is not human at the onset of any conversation.
“For sensitive interactions, people should know if they are dealing with a nonhuman,” Busse said.
The bill’s sponsor, Sen. Kirk Cullimore, R-Sandy, recognized that Utah is generally wary of increasing regulation for business owners; but deceptive business practices are already illegal. This bill, he said, just clarifies liability when AI is involved.
Business owners, executives and representatives — even those directly involved in AI software — told lawmakers they support such regulation.
Bree Jones, who owns two software companies, said AI is not some niche segment of the tech industry — it is “the predominant technology.”
Jones added, “I’m also a mother of two, and I have concerns about the fact that AI is becoming so dominant, so easy to access and build out, that those who are unregulated and maybe a bit mischievous with it could really do some harm here in the next short while.”
Ginger Chinn, vice president of the Salt Lake Chamber of Commerce, said the bill was one of the chamber’s top legislative priorities, because it encourages innovation and protects businesses.
Cullimore added that artificial intelligence is growing quickly, and it’s in the state’s best interest to stay ahead of it as best it can. The bill’s AI Learning Laboratory could allow businesses to test AI’s capabilities without fear of retaliation.
“We in Utah pride ourselves on having a light touch in industry and innovation, and we want to promote innovation,” Cullimore said.
Lab participants would, in the bill’s language, “analyze and research the risks, benefits, impacts and policy implications” of new AI technology and use their shared knowledge to help guide regulation. The participants would “define the things [they] want to learn, and decide what we need to regulate,” Busse said.
In other words, the lab would function a lot like generative AI does now.
Shannon Sollitt is a Report for America corps member covering business accountability and sustainability for The Salt Lake Tribune. Your donation to match our RFA grant helps keep her writing stories like this one; please consider making a tax-deductible gift of any amount today by clicking here.
Follow @ShannonSollitt