The emergence of online marketplaces has facilitated adaptive data collection but has also led to the need of adaptive decision making to counter non-stochastic patterns in the data. On the other hand, ethical concerns such as how to avoid discrimination have been only studied assuming that the input is i.i.d. In this paper, we explore the interplay between adaptivity and non-discrimination by studying the design of online learning algorithms that, when run on members of different groups, do not discriminate against some group. We consider the most basic question in such a setting: how can...ÂRead More