This article studies the use of AI in breast cancer detection in a real-world setting, comparing double reading using an AI-supported medical device against double reading without AI support. The data was collected from asymptomatic women aged 50-69 years who were part of Germany’s organized breast cancer screening program. One group of participants used the AI system for screening while the other group of participants had their mammogram results read independently by two radiologists and the breast cancer diagnosis results were compared for both groups. From this study, the authors found that across both groups 41.9 per 1,000 women were recalled for further assessment based on the test results. Furthermore, in the AI group 59.4% of the results came back normal with the rest coming back not normal (40.6%) compared to 53.3% normal for the control group and 46.7% not normal. The breast cancer detection rate (BCDR) per 1,000 women screened was 6.70 for the AI group and 5.70 for the control group, meaning the AI system detected one additional cancer per 1,000 screened women when taking the absolute difference. Lastly, the AI group had a lower recall rate (37.4 per 1,000) than the control group (38.3 per 1,000) but had an 8.2% higher biopsy rate than the control group. Overall, this study incorporates data from over 460,000 women, 119 radiologists, 5 different machine vendors and 12 screening sites across Germany to assess the use of AI in breast cancer screening. The findings from this study demonstrate that the use of AI in breast cancer screening results in higher BCDRs along with comparable recall rates to the control group.