Li metal is regarded as an ideal anode for the next-generation secondary batteries. However, the growth of Li-dendrite results in a low coulomb efficiency, thus restricting the commercial application of Li secondary batteries. The mechanism of Li deposition and growth in an atomic scale is still unclear, and there are different opinions about the origin of Li dendrite. Recent work has dealt with the application of machine learning in computational materials science. This review represented the applications of machine learning potential for the study of Li metal anode.