Introduction: Heart failure (HF) significantly affects public health by reducing quality of life and worsening prognosis. Early diagnosis remains challenging, as HF is frequently identified only when advanced symptoms appear. Each year, numerous patients undergo cardiac computed tomography angiography (CCTA) to assess suspected coronary artery disease (CAD). Due to overlapping risk factors and clinical presentations of HF and CAD, opportunistic evaluation for markers of HF during routine CCTA could facilitate earlier detection. Increased cardiac chamber volume, particularly left ventricular (LV) volume, serves as a promising marker for HF risk stratification. TotalSegmentator's heartchamber_highres module is an open-source artificial intelligence tool designed for automated segmentation of cardiac structures from CT imaging. This study aims to validate the accuracy of this module for estimating LV volumes from CCTA, using manually segmented cardiac magnetic resonance imaging (MRI) volumes as the reference standard. Methods: We analyzed 131 coronary CCTA scans from the Danish Dan-NICAD studies, comprising approximately 5,000 patients evaluated for suspected CAD. Of these, a subset underwent concurrent cardiac MRI, and 131 patients with both imaging modalities available were included. LV volumes were automatically quantified from CCTA images using the heartchamber_highres module and compared with manually segmented end-diastolic LV volumes from cardiac MRI. Correlation between automated CCTA-derived and MRI-derived LV volumes was assessed using linear regression analysis. Results: CCTA-derived LV volumes averaged 128.9 ± 29.8, which was higher but not significantly different compared to MRI-derived end-diastolic volumes of 125.3 ± 32.9 (p = 0.1369). The automated CCTA measurements showed significant correlation with the MRI reference volumes (R² = 0.41, r = 0.64, p < 0.001). Conclusion: Automated estimation of LV volumes from CCTA using TotalSegmentator's heartchamber_highres module demonstrates moderate to strong correlation with MRI-based LV volumes, supporting its potential utility for HF risk stratification in clinical practice.