We automatically annotate privacy policies using natural language processing and machine learning models trained using data from the OPP-115 Privacy Policy Corpus. Our annotations distinguish between ten categories of data collection and use practices.