Google Research published a prospective multi-center study demonstrating that an AI-powered ultrasound system can match expert sonographers in assessing gestational age, a critical measure for prenatal care. The research, published in JAMA Network Open, shows the system achieved a mean absolute error of 4.18 days compared to 4.48 days for expert sonographers when tested across diverse clinical settings in Chicago and Nairobi. The breakthrough addresses a critical global health gap: the World Health Organization estimates more than 250,000 preventable maternal deaths occur annually due to pregnancy complications, with two-thirds of the global population lacking access to diagnostic ultrasound. Google's approach uses "blind sweep" scanning—a standardized abdominal scan pattern requiring no specialized sonography training—combined with AI interpretation. The system was successfully deployed with novice operators using low-cost handheld probes, requiring only minimal training. The study's significance lies in its real-world validation across disparate settings. Originally trained on data from North Carolina and Zambia, the AI model required only minimal fine-tuning on Chicago data before prospectively testing in both locations, with no adaptation required for Nairobi. This demonstrates the system's generalizability across different patient populations, clinical workflows, and ultrasound hardware—a critical requirement for safe global scaling of medical AI.