Four U.S. elected officials, two Democrats and two Republicans, demand that a health care professional sign any medical decision and that the use of AI be disclosed to patients.
Tabled on September 1, 2026 by Representatives Tom Barrett (Michigan), Greg Landsman (Ohio), Buddy Carter (Georgia) and Kim Schrier (Washington State), the Doctors Not AI Act intends to circumscribe the role of automated systems in the coverage decisions of American health insurers. The text states that AI can only intervene in support: any assessment of medical necessity must be based on an independent clinical judgment rendered by a qualified professional, and a refusal of reimbursement based on this ground must bear its signature.
Two other provisions complete the system. Insurers would have to add the results produced by AI to the patient's administrative file and inform them of the use of such systems, which conditions the very possibility of challenging a decision. Finally, the text focuses on mental health, seeking to prevent mental disorders and addictions from being the subject of a more systematic algorithmic examination than other pathologies — a concern that is in line with the American rules of parity in mental health.
The tabling of this bipartisan proposal is part of a broader legislative movement: several states have already adopted their own laws governing the use of AI by health insurers in 2026.
Read the full article on the original source
Go to the source →