Healthcare executives said the White House's new principles are a pivotal step toward steering the application of artificial intelligence in the healthcare sector, emphasizing the need to ensure the technology's deployment is safe, equitable and responsible.
The AI principles, released by the White House on Dec. 14, have filled a void in the industry by establishing defining standards for the use of AI tools in healthcare. These principles, endorsed by 28 healthcare organizations, introduce a set of voluntary pledges aimed at guiding the implementation of cutting-edge AI models in healthcare delivery and payment.
Sanford Health's president and CEO Bill Gassen, an organization who signed the pledge, told Becker's that these principles are fundamental to the future of AI in healthcare.
"We applaud the efforts to convene a diverse group of healthcare organizations to coalesce around landmark voluntary commitments that will be fundamental to the future of AI and allow us to responsibly advance the use of these technologies for the benefit of those we serve," he said. "It has been energizing to collaborate over the last several weeks with colleagues across the healthcare ecosystem on a framework that reflects our shared commitment to harnessing large-scale AI and machine learning models safely, securely and transparently."
Highlighting the broader context, Mr. Gassen underscored the effect of President Joe Biden's executive order on AI, emphasizing the potential of these technologies to enhance various facets of healthcare, including quality, accessibility, affordability, equity, patient experience, clinician well-being and industry sustainability.
Another key player in this movement is Durham, N.C.-based Duke Health, whose CEO, Craig Albanese, MD, sees the pledge as a public affirmation of the health system's dedication to utilizing AI for the greater good.
Going beyond endorsement, Duke Health has taken an active role as a founding member of the Coalition for Health AI (CHAI). This coalition aims to establish guidelines and safeguards for the equitable and credible implementation of AI in healthcare. Duke Health has also committed to deploying equitable AI through developing a governance and evaluation framework for clinical algorithms used across the organization.
The Algorithm-Based Clinical Decision Support framework at Duke, which has been in place for more than three years, is now expanding to include frontier models. Michael Pencina, PhD, the inaugural chief data scientist of the health system, shared insights with Becker's into this expansion.
"With the framework in place, we are actively developing and testing multiple AI solutions aimed at reducing clinician burden, improving resource allocation, optimizing quality reporting and better predicting adverse health outcomes," he said.