In this session, the intersection of (official) statistics and machine learning will be explored. The emphasis will be on developing robust methodologies for leveraging machine learning within the framework of statistical practices. During the talks, a particular emphasis will be placed on the application of established statistical principles to ensure the accuracy, reliability, and generalizability of machine learning models that are used to generate official statistics. We will examine how traditional survey methodology concepts such as representativeness, sampling bias, and measurement error can be translated into machine learning models and addressed accordingly. Several frameworks will be discussed to ensure that machine learning is of high quality and methodologically rigorous, as well as address potential pitfalls, such as base rate fallacy, inherent to the data itself, that could be important when utilizing machine learning in statistical applications.