Dominik Herrmann demonstrates that operators of nameservers, which are used on the internet to resolve domain names into IP addresses, can track their users' behavior in more detail than previously thought. In particular, they can employ machine learning techniques to recognize individual internet users by their distinctive behavior and monitor them unnoticed over long periods. However, established methods are not suitable for anonymizing name resolution. Therefore, the author proposes new techniques for self-data protection and provides concrete recommendations for action.