The decisive factor for the use of machine learning (ML) in statistical data collection and evaluation is its positive influence on efficiency and the resulting flexibility in terms of the time required to compile statistics (Dumpert and Beck, AStA Wirtschafts- und Sozialstatistisches Archiv 2(11):83–106, 2017). The following chapter deals with the legal implications of the use of ML in official statistics, particularly with regard to data protection and the Artificial Intelligence (AI) Act that was recently passed (The European legislation referenced in this chapter applies to all statistical offices in the member states of the European Union. In some sections, the regulations applicable at the national level in Germany are also discussed. Analyzing the extent to which other national regulations influence the use of machine learning in official statistics goes beyond the scope of this chapter.). It finds, first, that data protection law is indispensable for official statistics but does not necessarily restrict the use of machine learning. Second, the recently passed European AI Act does apply to the use of ML in official statistics. However, it comes with only minor implications as the use cases would classify as low risk under the regulation.