Alibaba DAMO Academy, in partnership with several institutions, has successfully created an AI model named MAOSS, specifically designed for fatty liver screening. The groundbreaking research supporting this model has been published in Nature Communications. Utilizing 'plain CT + AI' technology, MAOSS marks a significant milestone by being the first model capable of simultaneously evaluating the severity of hepatic steatosis and staging liver fibrosis through non-contrast CT scans. The MAOSS model excels in accurately screening and classifying fatty liver, boasting an Area Under the Curve (AUC) that surpasses that of radiologists. Its implementation notably enhances the AUC achieved by doctors. It is adept at identifying patients at a heightened risk of stage 2 fibrosis, more than doubling the detection rate while preserving a high negative predictive value. Moreover, the model effectively forecasts the progression of liver cirrhosis. By relying solely on plain CT images, MAOSS eliminates the need for extra expenses for patients, minimizes the chances of missed diagnoses in high-risk fatty liver scenarios, advocates for an 'early intervention' strategy in managing chronic liver diseases, and enables timely high-risk alerts for patients at primary care centers or health screening venues.