Tuberculosis is an infectious disease caused by the bacterium Mycobacterium tuberculosis and represents a major global health challenge. Despite years of control, it has revived in recent years. China has the second highest incidence in the world, with a prevalence of 459 cases per 100,000 individuals aged 15 years. Radiography and chest pathology are the basic tools for detection and diagnosis. The small size and low number of tubercle bacilli make them difficult to identify under the microscope, leading to low detection rates and false diagnoses. Artificial intelligence (AI) has emerged as a promising tool for improving the accuracy and sensitivity of tuberculosis detection. The review summarizes the use of machine learning models for the automatic detection of tubercle bacilli and highlights the advantages of their integration into diagnostics. Understanding the onset and progression of tuberculosis is key to developing strategies for its diagnosis, treatment, and prevention.