Hmm. Sounds like great technology - but how exactly are they going to tell that your Kinect is pointing at the same room in which people are watching video, as opposed to, say:
1) A towel draped over the aforementioned Kinect?
2) The wall?
3) The hallway outside the room where the movie is being played?
4) Exactly half of the room, with the people who would bring your movie to an Unauthorised Number Of Viewers sat in the other half?
They would require you to set up authorized viewer's faces beforehand I imagine.
I see this as being years away at best though. They would have to build infrastruture around the movie delivery system for it to work - you'd have to register viewers faces and demand they set up the room correctly(well lit?), and of course, people would have to be cool with paying for that overhead. A competing service could easily continue doing things the way they do today, and thus get more customers, more money, and in turn pay content companies more. If it makes content companies more money, then I don't see them adopting the facial recognition model.
1) A towel draped over the aforementioned Kinect? 2) The wall? 3) The hallway outside the room where the movie is being played? 4) Exactly half of the room, with the people who would bring your movie to an Unauthorised Number Of Viewers sat in the other half?