In the United States, an artificial intelligence system has just received authorization to proceed with a task that previously required a doctor's involvement: prescribing medication for a patient. The state of Utah has authorized a pilot program by Nolla Health that will allow for the use of an AI system capable of evaluating acne patients and independently selecting the most appropriate treatment.

The process begins on a mobile phone: the patient—who must be an adult—completes a questionnaire taking approximately 10 to 15 minutes, providing information about their medical history and condition. They then use the smartphone to take detailed photos of their face. The artificial intelligence analyzes this information, assesses the condition, and selects a treatment from a specific list of medications previously defined by physicians.

The goal of the experiment is to reach a rather unusual scenario: an AI analyzing a patient, selecting a treatment, and issuing an initial prescription without waiting for a doctor's approval. That is precisely why the Utah experiment may be more significant than the acne treatment itself suggests.

What is being tested is not merely whether an artificial intelligence can recognize a skin condition, but rather how far we are willing to let AI go in the medical decision-making process before human intervention becomes necessary. Currently, the scope is limited to a specific list of acne treatments, but if this model proves safe and effective, the next question practically asks itself: what will be the next medical decision we are willing to entrust to a machine?