AIMS: To compare glucagon-like peptide-1 receptor agonists (GLP-1RA) with sodium-glucose cotransporter-2 inhibitors (SGLT-2i) for cardiovascular, atrial fibrillation (AF) disease course, and safety outcomes in adults with concurrent AF and type 2 diabetes. METHODS: Retrospective propensity-score-matched cohort study using the TriNetX Research Network (2016 - 2024). A new-user, active-comparator design identified adults with concurrent AF and type 2 diabetes initiating GLP-1RA or SGLT-2i. Patients were 1:1 matched on demographic, clinical, laboratory, and medication covariates. Hazard ratios (HRs) were estimated using Cox proportional hazards regression. The Benjamini-Hochberg procedure controlled the false discovery rate, and E-values quantified robustness to unmeasured confounding. RESULTS: After matching, 18,035 GLP-1RA users were compared with 18,035 SGLT-2i users (mean age 67.4 years; 52.2% male, 47.8% female). At 365 days, GLP-1RA was associated with lower all-cause mortality (HR 0.64, 95% CI 0.57 - 0.71), hospitalization (0.88, 0.84 - 0.92), 3-point MACE (0.78, 0.71 - 0.86), AF progression (0.94, 0.90 - 0.98), AF ablation (0.81, 0.70 - 0.94), and cardioversion (0.79, 0.70 - 0.90). Effects were consistent across age and body mass index subgroups. E-values ranged 1.5 - 2.5. CONCLUSION: Over 1 year, GLP-1RA initiation was associated with lower risks of mortality, hospitalization, atherosclerotic events, and adverse AF disease-course outcomes than SGLT-2i in adults with concurrent AF and T2D, with the established SGLT-2i benefits for heart failure and CKD remaining the basis for class choice in those subpopulations.