A closed-door workshop hosted by Stanford's Institute for Human-Centered AI brought together experts to discuss policy gaps and regulatory challenges for AI in healthcare.
The Stanford Institute for Human-Centered AI (HAI) recently convened a workshop of 55 leading experts ,to address the urgent need for updated regulatory frameworks for AI in healthcare. The meeting, hosted by HAI's Healthcare AI Policy Steering Committee, focused on three key areas: AI software for clinical decision support, healthcare enterprise AI tools, and patient-facing AI applications.
Participants highlighted the outdated nature of current regulatory frameworks, which were designed for an era of physical devices and analog data. They emphasised the need for substantial changes to effectively govern AI in healthcare, likening the current situation to driving a 1976 Chevy Impala on 2024 roads.
The workshop discussions covered a range of critical issues, including the need to streamline the FDA device clearance process for AI-powered medical devices with multiple diagnostic capabilities. Participants also debated the balance between human oversight in clinical AI tools, and the potential for increased administrative burden on healthcare providers.
Another key topic was the regulatory status of patient-facing AI applications, such as mental health chatbots. The experts grappled with whether these tools should be governed more like medical devices, or medical professionals, with a majority favouring governance similar to medical professionals.
The importance of including patient perspectives in the development and regulation of healthcare AI applications was a recurring theme. Participants stressed that this inclusion is crucial to ensure trustworthiness and address potential health disparities exacerbated by AI.
The workshop highlighted the complex challenges in regulating AI in healthcare, and the need for continued multidisciplinary research and discussions. HAI and its Healthcare AI Policy Steering Committee plan to conduct further research to support the development of policy frameworks that ensure safe, equitable, and effective use of AI in healthcare.