E-News Exclusive

Utah's New AI Law: Impacts on Radiology Software and Health Care Innovation

By Deeksha Shukla

The rapid advancement of AI in health care has opened up many prospects, particularly in the field of radiology. AI-driven software solutions are increasingly being integrated into radiology practices to enhance diagnostic accuracy, improve patient care, and streamline workflows. However, the fast-paced nature of AI innovation poses challenges for regulatory bodies, prompting different states to enact new laws to manage AI's integration into health care. Utah's new AI law represents a critical step in this evolving landscape, with significant implications for the use of AI in radiology.

Current AI Solutions in Radiology
As of May 13, 2024, the FDA authorized 882 AI and machine learning (ML)-enabled medical devices. These devices encompass a variety of tools designed to assist radiologists in interpreting medical images, identifying anomalies, and improving diagnostic accuracy. Despite the availability of these advanced tools, the regulatory framework governing their use is still catching up with the technological advancements. The FDA has been actively working to adapt its regulatory processes to better accommodate the unique challenges posed by AI and ML technologies.

Utah's New AI Law and Its Implications
On March 13, 2024, Utah’s law, the Utah Artificial Intelligence Policy Act (UAIPA), was signed and took effect on May 1, 2024. The UAIPA imposes disclosure requirements on entities using generative AI tools with their customers, limits an entity’s ability to blame generative AI for statements that violate consumer protection laws, and creates the Office of Artificial Intelligence Policy to administer a state AI program. The law requires health care providers to conduct thorough evaluations of AI software before integration into clinical practice, ensuring that these tools meet stringent safety and efficacy standards. Additionally, the law emphasizes transparency, mandating that patients be informed when AI is used in their care and providing them with the option to opt out.

For radiology, this law means that AI solutions must undergo rigorous testing and validation processes before they can be utilized. Radiologists and health care providers will need to ensure that their AI tools are compliant with both state and federal regulations, potentially slowing the adoption of new technologies but ultimately aiming to safeguard patient care. There are strict disclosure requirements and companies are responsible for the output of generative AI.

State AI Regulations
Several states besides Utah are now enacting their own AI regulations, highlighting the importance of state-level oversight in the absence of comprehensive federal laws. For instance, California has introduced legislation emphasizing ethical AI use, data privacy, and transparent decision-making. New York mandates regular audits of AI health care systems to meet both state and federal criteria, while Texas requires independent third-party evaluations of AI systems in health care to ensure their safety and efficacy. These diverse state efforts create a complex regulatory landscape that AI developers and health care providers must navigate.

Federal AI Regulations
At the federal level, ongoing discussions focus on the necessity of comprehensive AI legislation. While the FDA has progressed in updating its health care AI regulations, broader federal laws addressing AI applications in different sectors are still being debated. Key initiatives include the Algorithmic Accountability Act, which proposes impact assessments for automated decision systems, including those in health care, and the National AI Initiative Act of 2020, aimed at coordinating AI research across federal agencies without currently offering specific regulatory directives for AI applications. The absence of a unified federal regulatory framework means that AI regulation in health care continues to evolve on a state-by-state basis, with the FDA providing overarching guidance for medical applications.

Industry Implications
The evolving regulatory framework from the FDA and various state-specific regulations have significant implications for the health care industry. Developers and health care providers need to closely collaborate with the FDA to stay updated on new requirements, focus on generating postmarket data to improve device development and monitoring, and build trust through transparency in development practices and data usage. Additionally, supporting and educating health care providers about AI and ML devices is crucial for gaining patient trust and ensuring smooth integration into clinical workflows.

Impact of Utah's AI Law
Utah's AI law is a proactive measure aimed at ensuring the safe and ethical use of AI in health care. However, its stringent requirements may also present challenges. For radiology business owners, the need for thorough evaluations and validations can be time-consuming and costly, potentially slowing down the adoption of beneficial AI technologies. Lawyers will need to navigate the complexities of this new legal landscape, advising clients on compliance and representing them in any legal disputes that arise. Liability for generative AI’s output will be an interesting area that courts will need to consider carefully.

Regulators will face the dual challenge of enforcing these new standards while not suppressing innovation. Balancing safety with the need for technological progress will require careful oversight and ongoing dialogue with stakeholders. The state government will need to allocate sufficient resources to support these regulatory efforts, including funding for enforcement and education initiatives.

For AI developers, Utah’s law emphasizes the importance of transparency, disclosure, and patient consent, which are crucial for building trust in AI technologies. However, the added regulatory burden could potentially slow innovation and increase development costs. Developers will need to be proactive in engaging with regulators and ensuring their products meet the new standards.

Conclusion
Utah's new AI law, alongside similar measures in other states, emphasizes a careful approach to integrating AI into health care. These laws prioritize patient safety, data privacy, and transparency, aiming to foster public trust in AI technologies and promote responsible medical practices. As regulatory frameworks evolve, cooperation among state and federal authorities, health care providers, and AI developers becomes crucial to harnessing the benefits of AI in radiology while mitigating potential risks. Efforts by the FDA and state-level initiatives are pivotal in achieving this balance, ultimately aiming to enhance patient care through innovative and secure AI solutions. The health care industry stands at the brink of a technological revolution driven by AI. Successfully navigating regulatory challenges, particularly under laws like Utah's, will be essential to ensure this revolution enhances patient outcomes and streamlines health care delivery.

— Deeksha Shukla is a lawyer, a certified ISO 13485 auditor, and an ISO 9001 lead auditor. She currently works as the director for compliance and legal for RamSoft Inc.