Utah Enacts Artificial Intelligence Policy Act (UAIP) on May 1, 2024

As the calendar flipped to March 13, 2024, Utah introduced a pioneering piece of legislation in the technology industry—the Utah Artificial Intelligence Policy Act (UAIP). With this regulatory stride, Utah positions itself at the forefront of managing the burgeoning field of artificial intelligence (AI). For organizations deploying generative AI applications, such as AI text generator or AI video generator, the UAIP embodies a set of compliance measures that aim to balance consumer protection with the promotion of AI innovation.

Set to take effect on May 1, 2024, the UAIP stipulates certain obligations for entities using generative AI while interfacing with customers. This niche of AI encompasses systems that create content with human-like qualities—think of chatbot conversations or AI images generator, producing artworks indistinguishable from those made by human artists. In contrast, AI utilities that make product suggestions based on user predilections fall outside the act’s remit.

For individuals in regulated professions—which include most licensed healthcare practitioners—the Act mandates “prominent” notifications to consumers whenever they are interacting with generative AI. Such alerts must be expressively vocal, prior to voice communications, or clearly stated in electronic messages before any written interaction. The parametric details of what constitutes “prominent” are not pinned down within the legislation; nonetheless, the burden laid upon these regulated occupations implies that a marginal mention in privacy policies or terms of service will not be adequate.

Moreover, businesses outside regulated professions but under the umbrella of Utah consumer protection laws must provide “clear and conspicuous” disclosures regarding the use of generative artificial intelligence generated images or other AI outputs upon direct consumer inquiry. The method of inquiry and disclosure remains unspecified; however, businesses should interpret this directive in light of legal precedents which demand tangible, forthright consumer communication over implication or inference.

Perhaps one of the more pivotal tenets of the UAIP is the explicit prohibition that disallows companies from deflecting blame onto generative AI for any statement or act that contravenes consumer protection laws. In essence, the Act reinforces the notion that organizations should bear the same accountability for AI-generated communication as they do for that from their human employees.

To enforce these regulations, the UAIP has vested powers within the Utah Division of Consumer Protection (UDCP), which can impose fines up to $2,500 for each infringement. The courts can also implement additional remedies, including injunctions and profit disgorgement, where breaches are identified. Moreover, the Utah Attorney General’s Office can institute penalties of up to $5,000 per violation for non-compliance with administrative or court orders.

Tailored to foster a climate of AI-driven growth, the UAIP spearheads the creation of an Office of Artificial Intelligence Policy, responsible for developing the Artificial Intelligence Learning Laboratory Program (AI Lab). This initiative serves as a regulatory incubator, allowing companies trial runs to refine their AI systems under the auspice of “regulatory mitigation,” which includes diminished penalties and grace periods before fines are levied.

Looking ahead to the May enactment, companies must take constructive steps to align their practices with the UAIP’s directives. For example, businesses utilizing AI tools like AI video generator in customer engagements should engineer a disclosure strategy that provides consumers with clear information up-front—whether through a conspicuous written statement preceding an AI-driven chat interface or other means of direct notice.

Additionally, enterprises not expressly governed by the UAIP might prudently consider integrating similar disclosure protocols. Embracing such transparency could engender trust and establish better communications with clientele, ultimately serving as sound business conduct regardless of regulatory compulsions.

As the latest ai news & ai tools increasingly penetrate consumer-oriented services, the UAIP serves as a crucial policy blueprint for ensuring that generative AI is deployed responsibly and transparently. The Act precludes companies from absolving themselves of accountability for AI-generated content, as evidenced by cases like the British Columbia Civil Resolution Tribunal’s ruling against Air Canada in February 2024 for misrepresentations made by their chatbot.

The UAIP establishes that adherence to transparency and accountability remains paramount in the deployment of generative AI. As we navigate these new digital territories, companies undertaking the design and utilization of AI systems must not only aim for high standards of accuracy and reliability but also lead with forthrightness in consumer interactions—a balance that the UAIP strives to achieve.