Hate speech—communication that attacks a person or a group on the basis of identity factors, such as gender, race, or religion—is one of the main digital threats to democracy. Hate speech has manifold, empirically evidenced consequences for targeted individuals and groups experiencing systematic discrimination and for social cohesion as a whole. Yet, while the upheaval of social media has put the concept in the spotlight, such attention has also structurally transformed its meaning, turning hate speech from a concept with clear defining properties into a family resemblance comprising all kinds of online abuse. This process is far from causing only academic issues. It also sidesteps historical oppression as a defining property and as the reason for which one is targeted by hate speech. Thus, the process has been belittling public animosity against historically oppressed groups, reducing hate speech merely to a matter of offensive language on social media. This chapter shows how and why this conceptual change has taken place and the consequences it unleashes. It specifically addresses the problems of concept stretching, concept shrinking, and the inflation of concepts. Finally, it concludes that such conceptual issues jeopardize the potential that digital media research on hate speech has to provide guidance to a broad range of social actors.