Healthcare executives push for federal AI policy framework that preempts state laws

Healthcare executives from over two dozen leading organizations are advocating for a new federal artificial intelligence policy framework that preempts state laws on the technology. 

The Healthcare Leadership Council—a forum for C-suite leaders across different industries—recently released the results of a new survey of its members. It contends that, despite growing momentum in AI, disparate policies across states lines are stunting progress and impeding healthcare’s potential to improve outcomes and cut costs.

Leaders with hospital systems (such as AdventHealth and Sanford Health), insurers (Cigna, UnitedHealthcare), manufacturers (Novartis, Pfizer) and tech companies (Amazon, Surescripts) are now urging lawmakers to implement changes that might spur greater AI adoption.

“AI can further revolutionize patient care and reduce provider burden, but only if policymakers and industry move in lockstep,” Maria Ghazal, president and CEO of the council, said in a Jan. 15 announcement. “We need a clear, forward-looking national standard that harmonizes regulations and builds trust across all constituency groups, especially patients.”

A total of 27 organizations from the Healthcare Leadership Council’s membership participated in the interviews. Compiled by HLC and management and consulting firm ZS, the report is based on 30-minute, structured interviews with executives from across the healthcare ecosystem. Other participating organizations included multispecialty physician firm Envision Healthcare, healthcare improvement company Premier Inc., and accreditation agency The Joint Commission. 

Based on these interviews, the report details three primary barriers to AI adoption, also providing potential solutions to address them: 

1. Governance and regulatory complexity: Inconsistent and complicated AI regulations are preventing providers from integrating this technology at scale. Possible remedies could include establishing centralized legislation, modernizing regulations, clarifying accountability across all stakeholders and ensuring liability protections for responsible use in patient care. 

2. Data access and infrastructure challenges: Issues related to information sharing and privacy may be preventing widespread adoption. Recommendations to address this could include strengthening national standards for data quality and interoperability, establishing federal benchmarks for clear disclosure of AI capabilities, reducing concerns about potential bias, and standardizing terminology. 

“For AI to be developed for purposes such as diagnosing, surfacing incidental radiology findings or recommending treatment protocols, training sets must be diverse, representative and comprehensive,” the report notes. “Enforceable data standards and accessible infrastructure are essential to realizing AI’s full potential.”

3. Capabilities and end-user trust: Skill gaps and trust concerns across healthcare may be inhibiting effective implementation of the technology. Executives recommend addressing this barrier by embedding AI literacy and ethics into K–12 education and upskilling healthcare professionals through AI-focused training. 

“This report serves as a practical guide for how public and private stakeholders can work together to unlock AI’s full potential in healthcare,” Bill Coyle, chairman of ZS, said in the announcement. “Removing these barriers can improve care quality, strengthen the healthcare workforce, and advance more patient-centered care across the healthcare system,” he added later. 

The report comes after the White House in December issued an executive order aiming to curb excessive overregulation of artificial intelligence at the state level.