Last week, YouTube announced on its company blog (in an entry titled “A YouTube for All of Us”) that it is tightening its restrictions on sexual content and profanity. Of course, YouTube has always had limits, mostly for pornography, spam, and gratuitous violence, handled primarily through automatic filtering that can spot X-rated scenes, and through the user community itself flagging inappropriate content for review. Now that user community is in an uproar about the recent announcement, because the restrictions will extend to sexually suggestive video and video that uses profanity. It’s not a surprise that sites like YouTube have to strike their own balance, between being an open platform for whatever users choose to post, and building a user community (not to mention a public brand) that’s acceptable to mainstream users and to the sponsors eager to sell to them. Censorship is hardly new to the Internet. What is new is the way YouTube intends to handle inappropriate videos: not only by removing some videos and placing age restrictions on others, but through “demotion.” “Videos that are considered sexually suggestive, or that contain profanity, will be algorithmically demoted on our ‘Most Viewed,’ ‘Top Favorited,’ and other browse pages.” This means that videos with too much profanity or sexually suggestive content will not be removed, but their popularity will be mathematically reduced, so they don’t show up on the lists of what’s most popular - censorship through technical invisibility. And we won’t know which videos, for what reasons. That YouTube can bury the rules, and their judgments, into the mechanisms by which users know what’s available and popular, points to the kinds of free speech dilemmas we’re likely to face in a digital future, and that we’re hardly prepared to think through.