Huh? Totally Clueless German Court Says ContentID Isn't Good Enough, YouTube Must Block Infringement By Keywords
from the total-failure dept
We’ve discussed a few times the legal fight in Germany between YouTube and the overbearing collections society GEMA. German law is a bit bizarre in that it has very little regard for secondary liability protections, and often seems to default to blaming third parties for the actions of their users. On top of that, GEMA is incredibly powerful and controlling in Germany. When I was there a few years ago (in part to discuss this case), I had musicians explaining to me that they had “secret websites” because GEMA wouldn’t let them offer their own music for free. The GEMA/YouTube dispute centers around the fact that GEMA wants YouTube to pay a fixed fee every time a video that includes a GEMA-covered song is played (and GEMA has actually suggested the fee for each stream should be identical to the cost of a download — no joke). After suing YouTube/Google, GEMA refused to negotiate, making Germany the only modern country in the world in which a collection society didn’t work out a deal with YouTube (and meaning that Google started blocking tons of music videos in Germany).
Unfortunately, the court has now ruled in the case, and the results seem ridiculous. It has said that YouTube is liable when users post videos (third party liability concepts still just don’t make sense to German courts). Even more ridiculous, however, is that the court has said that YouTube’s famous ContentID system is not enough. Instead, it must also install a keyword-based filter to block GEMA songs from being uploaded. Keyword-filters? Really. We’ve done this before a bunch of times and it doesn’t work. At all. Keyword filters are really stupid ways to deal with these kinds of things. First off, they tend to block all sorts of legitimate content. But, more importantly, users figure out how to get around them in less than an hour. They just start coming up with easy-to-decipher substitutes. Comparing a keyword filter to ContentID is like comparing a human strapping on wings to a modern fighter plane. One of them works and can actually get the job done. One of them just makes you look like an idiot.
The only “concession” the court appears to have given YouTube is that it only expects such filtering to work going forward, rather than having them search the archive. That, of course, is barely a concession at all. If I remember correctly, this particular court, in Hamburg, is somewhat notorious for siding with copyright holders, so I wouldn’t be surprised to see an appeal on this case…