Delimiting the contours of hate speech is not easy, especially when the scope of exploration is the web, with its blurred boundaries. With the advent of the Internet and the development of social media, the idea of public discussion has changed, no longer populated only by subjects institutionally responsible for the production of information, but also, and increasingly pervasively, by individual, non-professional subjects, who constitute a galaxy of “informal” information sources. In the diffusion of hostile rhetoric, the responsibilities are, therefore, multiple, transversal and not always easily identifiable and punishable. The recourse to the web as a vector of incitement to hatred still today raises unprecedented questions and imposes the search for adequate answers at the legal, political, institutional, social, cultural and media levels. These responses unfortunately cannot keep up with the rapid mutations of the phenomenon. However, some recent events show that sometimes the xenophobic and racist mud machine can get jammed up in front of alternative narratives that are not subordinate, autonomous, creative and intelligent. The contribution tries to give a picture of this evolution mentioning some exemplary cases.