If a prior probability is given as part of the setup of the problem, then use that information (i.e., use MAP). If no such prior information is given or assumed, then MAP is not possible, and ML is a reasonable approach.