FDA’s Digital Health Center of Excellence director Rick Abramson said GenAI-enabled tools will likely need to be assessed “like a physician,” placing algorithmic behavior inside a clinical accountability framework. The statement implies that regulators may focus less on device analogies and more on performance under clinical judgment conditions. The comment comes as the FDA and healthcare ecosystems grapple with how to oversee models whose outputs may evolve with changing inputs, data distributions, and user workflows. For developers, the signal is that clinical validation and operational controls may carry even more weight when models act in decision-support roles. Stakeholders in clinical AI should treat the physician-comparison as a guardrail: documentation, monitoring, and evidence generation methods that mimic clinical evaluation pathways may be expected. The statement also heightens urgency for robust post-market evaluation plans.