With the rise of digital systems in learning scenarios in recent years as learning management systems, massive open online courses, serious games, and the use of sensors and IoT devices huge amounts of personal data are generated. In the context of learning analytics, this data is used to individualize contents and exercises, predict success or dropout. Based on a meta analysis it is investigated to which extent the privacy of learners is respected. Our research found that, although surveys have shown that privacy is a concern for learners and critical to adopt to establish trust in learning analytic solutions, privacy issues are very rarely addressed in actual learning analytic setups.