In patients in the intensive care unit (ICU), acute respiratory distress syndrome (ARDS) is a frequent pulmonary condition that can be fatal. It is often linked to direct lung damage, which is followed by sepsis and multi-organ system failure. According to estimates, 10% of all ICU patients, 23% of ventilated patients, and 5.5 cases per ICU bed experience ARDS each year in critical care. ARDS patients have an overall death rate of 35–50%, which accounts for 4% of all hospital admissions and hasn't altered in ten years despite improvements in the healthcare system. In this paper we have done a comparative analysis of AI-driven approaches have gained significant traction in the diagnosis, prediction, and management of ARDS. This analysis compares various methodologies and AI applications across key studies in this field. Our main focus is on different machine learning (ML) models used for prediction purposes.