As AI becomes more involved in prior authorization decisions, who is ultimately accountable when something goes wrong? Michelle Mello discusses how existing safeguards work, where gaps remain, and how AI could help make the appeals process less burdensome for patients and clinicians.