Artificial intelligence in cancer screening: a narrative review of current evidence and future directions

Cancer remains one of the leading causes of death worldwide, with outcomes improving significantly when the disease is detected at an early stage. Screening programs have demonstrated clear benefits, yet they often struggle with limited accuracy, unequal participation, and substantial demands on health care systems. Artificial intelligence (AI) has entered this space as a promising tool, aiming to improve detection, reduce workload, and support more individualized screening approaches.

This review explores the application of AI across different aspects of cancer screening. We discuss its use in established imaging modalities such as mammography and low-dose CT, explore early developments in liquid and blood-based assays, and highlight efforts to harness routine clinical and laboratory data to better identify people at risk.

AI is beginning to move beyond experimental settings and into real-world clinical practice. Early evidence suggests that it can enhance both accuracy and efficiency. At the same time, important challenges remain, including risks of bias, overdiagnosis, and uneven performance across clinical environments. However, sustainable progress will require robust validation, careful integration into clinical workflows, and policies that ensure the technology benefits patients across diverse settings.