Dive Brief:
- The White House on Thursday announced a voluntary commitment from 28 healthcare organizations, including CVS Health, Geisinger and nonprofit health system Allina Health, to use and buy “safe, secure, and trustworthy” artificial intelligence products.
- The payers and providers promised to inform patients whenever they receive content that was generated by AI and not reviewed or edited by a person. The signees also agreed to follow a risk management framework when using applications powered by foundation models, or systems built on broad datasets that can be used for a number of general tasks.
- The pledge comes as interest in the technology has spiked over the past year, particularly in generative AI, which can create new text or images.
Dive Insight:
The White House commitment asks payers and providers to develop and investigate tools that advance health equity, expand access, cut costs, improve care coordination and reduce clinician burnout.
AI has the potential to positively impact the sector, helping to quickly analyze imaging results, develop new drugs, and interpret and utilize vast amounts of patient data, the Biden administration said.
A survey conducted earlier this year by the American Medical Association found 65% of physicians saw an advantage to AI in healthcare, with the greatest enthusiasm for tools that could reduce administrative burdens in medical documentation or prior authorization.
But some experts and researchers have raised concerns about a rapid deployment of the technology, arguing inaccuracy and bias add significant risks.
Insurers have also faced legal challenges over the use of AI in appraising claims. Earlier this week, Humana was sued over an algorithm that allegedly prematurely cut payments for Medicare Advantage beneficiaries’ rehabilitative care, pushing patients to pay out of pocket or forgo care.
Regulators and lawmakers have shown interest in guardrails too. The HHS’ Office of the National Coordinator for Health Information Technology finalized a rule this week that includes transparency requirements for clinical decision support and predictive tools certified by the agency.
Some users will be able to access information like the value of the intervention, how it should be used, known risks or inappropriate uses, and how the tool is maintained and updated.
The White House signatories include health systems Mass General Brigham, Boston Children’s Hospital, Sanford Health and Hackensack Meridian Health.