I absolutely agree and we did our best to automate as much of the process as possible. There were a few problems that required a pair of human eyes though. First, no copyright holder sent a complaint in the same format. Some sent a simple email with a file name and ip/time stamp while others sent a long poorly rendered xml file that was automatically sent from some contractor's hacked torrent client. There was no rhyme or reason to the manner in which we received the data. Another issue was file name problems. The copyright holder would send out complaints for a user sharing the movie "Dune" when in reality the file was only a copy of pictures taken dirt biking at the local dunes.

The automation mostly came in pulling known data like our IP block from the complaints, and having a good IP logging system. We broke the process down to 3 steps, each very quick. A complaint only took 3-5 seconds to process, so the issue was just the sheer volume in them.

When the end result is the possibility of a lawsuit, it is in everyone's best interest to maintain a human element in the chain.