The year 2023 experienced record-breaking marine heatwaves (MHWs) across the North Atlantic and Mediterranean Sea, contributing to the highest global surface air and sea surface temperatures (SSTs) on record. These events were exceptional in intensity, persistence, and spatial extent, reflecting the combined influence of anthropogenic warming, short-term climate modes and complex atmosphere-ocean interactions. This study investigates the large-scale atmospheric drivers behind these extremes using NOAA OISST v2.1 and ERA5 reanalysis datasets. We characterize MHWs from May to August 2023, analyze surface and mid-tropospheric anomalies in pressure, air temperature, wind and airāsea heat flux and apply regularized generalized canonical correlation analysis (RGCCA) to study multivariate links between atmospheric variability and MHW characteristics. Our findings show that the Subtropical Atlantic experienced the longest MHW, the Northwest Atlantic the most intense, and the Western Mediterranean the most frequent events. The summer North Atlantic Oscillation (NAO) and Scandinavian Pattern (SCAN) emerged as key modulators of MHWs. Compound configurations of NAOā/SCAN+ in July-August and NAO+/SCANā in May-June generated persistent atmospheric ridges and weakened the Azores High, which in turn suppressed winds, altered heat fluxes and mixed layer depths, and promoted stratificationāleading to sustained surface warming. The leading RGCCA mode explains more than 40% of the SSTA variability and shows statistically robust correlations (r = 0.81ā0.94) between atmospheric drivers and MHW evolution. This multivariate approach demonstrates how teleconnection patterns co-modulate regional MHW dynamics, underscoring the importance of compound atmospheric influences. Our results highlight the utility of RGCCA in diagnosing complex climate extremes and support the integration of large-scale atmospheric indicators into early warning systems and adaptation planning in the face of increased marine heat stress.