Oral potentially malignant disorders (OPMDs) are a group of oral mucosal diseases with different morphological characteristics that have an increased risk of progressing to oral squamous cell carcinoma (OSCC). Epidemiological studies indicate that approximately 4.67% of the global population may have an OPMD, and meta-analysis results show that the risk of progression to OSCC in patients with OPMD is 2.5 times that of the general population. The progression of an OPMD to OSCC is a complex process driven by risk factors including smoking, alcohol consumption, betel quid chewing, and oral microbiota. It is characterized by intricate interactions among genetic alterations, epigenetic modifications, and dysregulation of the tumor microenvironment. At the molecular level, the intrinsic basis of epithelial carcinogenesis is established by genomic loss of heterozygosity and cascading dysregulation of critical signaling pathways, while crucial external support for malignant transformation is provided by the immunosuppressive microenvi-ronment, including M2 macrophage polarization, immune evasion mediated by programmed death-ligand 1, and the protumorigenic effects of oral microbiota. Although a variety of therapeutic modalities (e.g., pharmacotherapy, surgical treatment, and photodynamic therapy) have been employed for the prevention of OPMD malignant transformation, no universally accepted treatment has been proven to completely eradicate the risk of cancer development at present. In recent years, artificial intelligence technologies have demonstrated promising performance in predicting the malignant transformation risk of OPMD by integrating multi-dimensional clinical, pathological, and molecular data, thereby providing new decision-support tools for advancing individualized risk stratification and precise follow-up management. This review summarizes recent advances in the prevention of OPMD malignant transformation, with the aim of offering a reference for the clinical management of OPMDs.