Internet censorship has evolved. In Version 1.0, censorship was impossible; in Version 2.0, it was a characteristic of repressive regimes; and in Version 3.0, it spread to democracies who desired to use technology to restrain unwanted information. Its latest iteration, Version 3.1, involves near-ubiquitous censorship by democratic and authoritarian countries alike. This Article argues that the new censorship model involves four changes: a shift in implementation to private parties; a hybrid approach mixing promotion of favored viewpoints with suppression of disfavored ones; a blend of formal mandates with informal pressures; and a framing of censorship using uncontroversial labels. It suggests a set of responses to censorship that cabin its abuses and push it towards more legitimate methods: focusing on governmental restrictions, insisting on labeling censorship as such, supporting distributed Internet governance, demanding a default right of access to information, and addressing corporate involvement.