Campaign 001 — Message
Type: Email Subject: The AI arms race in utilization review — where does it end?
Hi [First Name],
Wanted to share a Health Affairs piece that's been on my mind — "The AI Arms Race in Health Insurance Utilization Review" from Stanford's Michelle Mello and team.
https://www.healthaffairs.org/doi/10.1377/hlthaff.2025.00897
84% of large insurers are now using AI operationally. Providers are also deploying AI-powered appeals at scale. Both sides are automating against each other — and Mello's data suggests neither side is governing it well. Over 25% of insurers don't test for accuracy or bias. Nearly 40% have no committee reviewing AI in UM decisions.
The part that stuck with me: if both sides just layer AI on top of the same adversarial process, we end up right back where we started — processing faster, but nobody actually making better decisions. The article frames a different possibility, where AI gets the initial decision right and clears routine approvals instantly. But that requires someone to build toward it.
I'd love to hear how you're thinking about this. Genuinely curious whether the people closest to these decisions see a path where AI changes the dynamic, not just speeds it up.
Happy to hop on a call if this is on your radar.
Michael Yuan Co-founder, DaisyAI