In response to rapidly-evolving technology, Utah has launched a new Office of Artificial Intelligence Policy.
State leaders said the office is aimed at staying on top of technology changes, fostering innovation, protecting citizens and building trust. The office was created by lawmakers earlier this year and will be housed in the Utah Department of Commerce.
Gov. Spencer Cox applauded the office during a launch event on July 8. He said it will be a collaborative effort between government and industry, and will balance innovation with protecting consumers. And because of that collaboration, if AI legislation needs to happen quickly, Cox said lawmakers can have a special session.
“I'm proud of the ‘Utah Way’ that encourages us to do this, that business and government can work side by side in a way that helps everyone and elevates our state and powerful way.”
The office will be led by Brigham Young University mathematics professor Zach Boyd, who is currently on leave to oversee this statewide project.
One of its main functions, Boyd explained, will be its Learning Laboratory. The lab will focus on understanding pressing AI issues and making policy recommendations to the state Legislature.
The first topic will be to examine generative AI in mental health care.
Margaret Busse, the executive director of the Utah Department of Commerce, said there’s a long list of topics the lab could and likely will focus on in the future, including the use of AI in K-12 education. But Busse said they are starting with mental health care because of the prevalence of mental health issues in Utah, the shortage of resources and because it addresses multiple AI issues, like data privacy and health care.
Within the next few months Busse said the office plans to make recommendations to the legislature regarding AI in mental health care. One of the questions the lab will be looking at is guardrails for what AI chatbots are and are not allowed to do in mental health care.
Another main function of the lab will be regulatory mitigation.
Boyd said some of what companies and individuals want to do with AI could be illegal because of laws written decades ago, or it could be unclear how it works with the state’s regulations. In these cases, companies and individuals can apply to the AI Office for “regulatory mitigation agreements.”
The Legislature has given the office the power to work with regulators to come up with agreements to temporarily deploy new AI products or technologies. Those agreements could include things like reduced fines for regulatory violations. Boyd said this will also give the state an opportunity to gather preliminary data about the technology.
The first to apply for regulatory relief is ElizaChat, a company aimed at providing mental health support for teenagers.
Boyd said there are potential benefits to providing support with AI, but also risks.
“The lab is currently working with ElizaChat to develop the right parameters for initial small scale rollout of these technologies, and the right gateways that you have to pass through in order to expand to larger populations as the technology may be able to prove itself.”
The state law that created the office also addressed consumer protections. It requires companies to disclose when someone is interacting with generative AI.
While talking with reporters after the event, Cox said another concern with AI is the amount of power that data centers use.
“We do not produce enough power as a country for the data centers right now.”
Cox added there was talk of calling a special session to address this. He said by the next legislative session in January, there will be legislation to encourage generating more power.
“We want it to be environmentally friendly power generation, obviously. And see if there aren't some more things we can do as a state to wait to get ahead of that power curve.”
