Artificial intelligence cannot explain its mistakes. Proponents say the proposed bill aims to protect consumers from deception by robots.
Generative artificial intelligence (the kind that generates text and images based on prompts) learns much of the same way humans do. With each interaction with a human, it grows its vocabulary, recognizes language, and stores those encounters as data so it can respond better. future.
Utah lawmakers want people in Utah's business sector and the state Department of Commerce to learn about AI as well.
The Senate Business and Labor Committee advanced SB149 to the full Senate on Thursday by a unanimous vote. If passed, the bill would create an AI “learning lab” for companies and state regulators to monitor the latest technologies and trends in artificial intelligence, so states can better understand how to regulate AI. It becomes like this.
The bill would impose some immediate rules on consumer AI. Because AI can already act like a human, but that doesn't make it human.
“We want to put guardrails in place now,” Margaret Busse, executive director of the Utah Department of Commerce, told the committee.
SB149 could subject companies that use AI to liability if the AI they generate misleads consumers in violation of Utah's consumer protection law. For example, if a chatbot misleads a consumer into purchasing a product, the chatbot is not responsible for the lie, but the company behind it.
“AI cannot be used as a defense,” Busse said.
The bill would also require consumer-facing generative AI, such as chatbots and text messages, to answer honestly when asked, “Are you a human?” AI in certain licensed industries, such as healthcare, mental health, and finance, will be required to reveal that they are not human at the beginning of a conversation.
“In sensitive interactions, people need to know whether they are dealing with a non-human,” Busse said.
The bill's sponsor, Sen. Kirk Cullimore (R-Sandy), acknowledged that Utah is generally wary of tightening regulations for business owners. However, deceptive business practices are already illegal. He said the bill only clarifies liability when AI is involved.
Business owners, executives, and representatives, including those directly involved in AI software, told lawmakers they support such regulations.
Bree Jones, who owns two software companies, said AI is not a niche area of the tech industry, but a “mainstream technology.”
Jones added, “I'm also a mother of two children, and I think that AI has become so dominant, so easy to access and build on, so unregulated, so perhaps even a little naughty with it.'' “We are concerned about the fact that people who have done so could be at real risk,” he added. Some harm will be done here soon. ”
Ginger Chin, vice president of the Salt Lake Chamber of Commerce, said the bill is one of the chamber's top legislative priorities because it fosters innovation and protects businesses.
Cullimore added that artificial intelligence is growing rapidly and it is in the state's best interest to stay ahead of it as much as possible. His AI Learning Laboratory in this bill could allow companies to test AI capabilities without fear of retaliation.
“At Utah State, we pride ourselves on being light on industry and innovation, and we want to foster innovation,” Cullimore said.
According to the bill's language, lab participants will “analyse and study the risks, benefits, impacts, and policy implications” of new AI technologies and use their shared knowledge to support regulatory guidance. Become. Participants “define things” [they] We want to learn and determine what needs to be regulated,” Busse said.
In other words, the lab will function much like today's generative AI.
Shannon Sollitto is Report for America Business Accountability and Sustainability Officer of the Salt Lake Tribune. Your gift, matched by an RFA grant, will help her keep writing stories like this one. Click to consider making a tax-deductible gift of any amount today. here.