Atrial fibrillation (AF) is a major cardiovascular disease with a prevalence of 1.7% of the total population in Korea, associated with 25% of ischemic stroke and 30% of heart failure, and is a major cardiovascular disease that doubles the risk of dementia. AF catheter ablation (AFCA) is an effective procedure that lowers the risk of heart failure mortality and cerebral infarction and improves cognitive or renal functions. However, the recurrence rate after the procedure is relatively high, especially in patients with long-standing persistent AF in which atrial remodeling has already progressed. Research on the prediction of treatment efficacy using artificial intelligence (AI) is being actively conducted around the world. We predicted the AFCA poor responders who will progress to permanent AF despite AFCA among a total of 3,372 patients included in the Yonsei AF Ablation cohort and the 2nd independent cohort with a long-term follow-up through AI with area under curve (AUC) 0.943. Therefore, in this prospective randomized clinical study, the difference between the patient selection for AFCA using AI algorithm and the clinical guidelines-based decision will be compared and evaluated in terms of long-term rhythm outcome.