Researchers developed ECG-ASCVD, an artificial intelligence system that analyzes routine electrocardiograms to predict atherosclerotic cardiovascular disease risk. Testing across 500,000+ individuals from Yale, Brazil's ELSA study, and UK Biobank, the AI achieved 68-75% discrimination accuracy (C-index 0.68-0.75) and remained predictive even after accounting for traditional risk scores like PREVENT. The system works with standard 12-lead ECGs, ECG images, or even single-lead recordings. This represents a significant advance in cardiovascular risk assessment because ECGs are widely available, inexpensive, and routinely performed in clinical settings worldwide. Unlike traditional risk calculators that require multiple lab values and clinical measurements often unavailable in primary care, this AI approach could identify high-risk patients using only a basic heart rhythm test. The technology could be particularly valuable in resource-limited settings where comprehensive risk assessment isn't feasible. However, this is a preprint awaiting peer review, so results may change. The moderate accuracy suggests this tool would complement rather than replace existing risk assessment methods. While promising for population-level screening, the clinical impact will depend on how well it performs in real-world deployment and whether it meaningfully improves patient outcomes.