Atrial fibrillation (AF) is a common cardiac arrhythmia that is associated with elevated thromboembolism risk caused by multiple pathophysiologies, including a hypercoagulable state, structural heart changes, left atrial appendage stasis, inflammation, and endothelial dysfunction. With the exception of lone AF, most other categories of AF, whether paroxysmal or persistent, have been shown to share a high thromboembolism risk. Risk stratification schemes such as CHADS2 and CHA2DS2-VASc scores help to identify the level at which anticoagulation may mitigate thromboembolism risk. AF may be episodic and asymptomatic; therefore, AF diagnosis that depends entirely on office electrocardiogram (ECG) may be easily missed. With the increasing use of pacemakers, implantable cardioverter defibrillators (ICDs), and insertable loop recorders (ILRs) for diagnosis and treatment of arrhythmias, AF has been incidentally detected with increasing frequency. However, the sensitivity and specificity for detection of AF, especially brief episodes, vary from one type of device to another, and rhythm confirmation should be considered. Several recent studies have examined device-detected AF and have tried to follow associated clinical outcomes. In this paper, we review studies that have addressed device-detected AF and associated thromboembolism risk to try to identify the burden of AF that is associated with an elevated risk of thromboembolism and may therefore warrant anticoagulation therapy.