A research team developed AI software designed to detect paroxysmal atrial fibrillation (PAF) from ECG traces recorded during sinus rhythm, addressing a key diagnostic blind spot for intermittent AF. The model analyzes subtle electrical patterns to flag patients at risk of undetected arrhythmia and could enable earlier anticoagulation decisions or targeted monitoring. Paroxysmal AF is an intermittent arrhythmia that can evade single‑timepoint ECGs; detecting it from sinus‑rhythm signals would expand screening reach without prolonged monitors. Clinical validation and integration with existing ECG workflows will determine the tool’s uptake in cardiology and primary‑care settings.