I talked with a copyright lawyer, Gregory Rutchik of The Arts and Technology Law Group in San Francisco, today about the Viacom lawsuit against Google and YouTube. His main question: "Why isn't every piece of content coming out watermarked?" He's referring to the notion of digital watermarking, a visible or invisible tag that carries information about the content.
With such watermarking, Google and YouTube could easily filter out uploaded videos carrying the watermark--unless, of course, they have an explicit deal with the content provider. In other words, Rutchik thinks there's no reason Google or any other site hosting these videos should be responsible for coming up with far more difficult filtering technology based on who knows what when there's an apparently obvious fix that the content owners themselves could implement.
But is that true? How far along is digital watermarking? I found an IBM Research article from a decade ago that indicated it was quite new, but are we still so far from an effective watermarking solution today? It sure doesn't look like it. In fact, here are some testimonials from Paramount Pictures and the MPAA. But I could be missing something. Seriously, is there any technological impediment to using watermarks?
Rutchik's answer is that media companies don't really want watermarking--one, because they realize they get a branding boost even from illicit uploading of clips, and two, because watermarking would force them to confront the notion that they can't expect viewers to flock to their site and only their site to see content they want. But the barn door's open, the horse long gone, isn't it? The Internet is a new distribution channel that won't be shut down with even a billion-dollar lawsuit.
Update: Maybe I just called this technology by the wrong name. Om mentions that there's likely to be a bull market soon in video fingerprinting companies.