The requirements on explainability imposed by European laws and their implications for machine learning (ML) models are not always clear. In that perspective, our research analyzes explanation obligations imposed for private and public decision-making, and how they can be implemented by machine learning techniques.