This startup is issuing AI-generated acne prescriptions

The news that Nolla Health is issuing AI-generated acne prescriptions in Utah is a classic Tech Pulse story: a supposedly niche feature that actually exposes the fault lines of an entire industry. On one hand, this is a practical application of computer vision and clinical decision support—acne is highly visible, well-documented, and relatively low-risk compared to other conditions, making it a sensible first target for autonomous prescribing. On the other hand, the framing that an AI system can "autonomously" prescribe a medication, with no clinician in the loop, deserves serious scrutiny. The core issue is accountability. When a dermatologist writes a prescription, they take responsibility for the diagnosis, the choice of drug, the dosage, and the monitoring plan. When an AI writes one, who owns the error? The startup? The model's training data? The patient who clicked "accept"? Nolla Health presumably has some oversight mechanism, but the Bloomberg report and the company's own announcement emphasize the autonomy of the system. That emphasis is a red flag, not because AI can't be useful in dermatology, but because prescription decisions involve judgment calls about side effects, contraindications, and patient history that a face scan alone cannot capture. There's also a regulatory tension here. The FDA has been wrestling with software as a medical device for years, and the rapid expansion of AI prescribing tools is pushing those boundaries faster than the rules can adapt. Utah may be a permissive environment for this kind of pilot, but what happens when a patient in another state has an adverse reaction? The legal infrastructure for AI-prescribed medication is nowhere near mature. Nolla's service is a test case for how the rest of the country—and the world—might handle AI-driven clinical autonomy. Still, it would be reductive to dismiss this as reckless. Acne is a quality-of-life condition that disproportionately affects young people, and access to dermatologists is often limited by cost, geography, and wait times. If Nolla's AI can safely triage and treat mild to moderate acne, that's a genuine improvement in access. The question is whether the company is being transparent about the limitations of its system, and whether it has built in enough human fallbacks to catch the cases where the AI is confidently wrong. That transparency will determine whether this is a step forward for healthcare or just another example of AI hype outpacing responsibility. Source: https://www.theverge.com/ai-artificial-intelligence/1005075/nolla-health-acne-ai-prescriptions