Text and Data Mining (TDM) is usually performed without proper analysis of the legal status of the databases/datasets that are to be TDMed. I am curios about how researchers address the issue (if at all) and whether the legal status of the original databases/datasets has an impact on how they license their results (if at all).