Where Copyright Enforcement Collides With Network Management

ArsTechnica brings us this story about the fascinating next step in the arms race between online filesharing and the content industries. A new P2P client, developed by the University of Washington, integrates BitTorrent into a social-networking-type framework that renders it impractical for any user to determine the ultimate origin of files he downloads.

The way this apparently works is that a user installs the client and links to his buddies (who have likewise linked to their buddies). When he runs a search it will poll the buddies to see if anyone has the desired file. If not, it will search the buddies' buddies, expanding outward until it finds the file. The file is then routed back to the original user through each intermediate buddy and, significantly, the source of the file is anonymized at each step so that each client is aware only of its immediate neighbors. Consequently an MPAA executive searching the network for infringing files would have no idea which users to sue. They would not have IP addresses or any other identifying information for anyone beyond their immediate circle of friends (supposing, for the sake of argument, that MPAA executives have friends).

It's a great system for people who want to avoid being sued for copyright infringement. The flip side, however, is that routing files through a bunch of intermediate steps rather than directly from the ultimate source to the downloader is hugely inefficient. You burn a lot of network resources to achieve anonymity. This is where the interests of the content industries knock heads with the interests of the ISPs. The ISPs have long had a love-hate relationship with filesharing as it is a network management challenge, but also a great value proposition for their customers. This is less true now, but in the early days filesharing was one of the primary drivers for consumer adoption of broadband. Not only would ISPs be better off with the straight BitTorrent model, but BitTorrent could be optimized to prefer nearby nodes, thereby increasing its efficiency and lowering the burden of filesharing on ISPs. This can be done to some extent by software alone, but could be further enhanced through the cooperation between the software developers and the ISPs. Moreover, the FCC's decision in the Comcast case makes clear that the ISPs are on dangerous ground when they attempt to interfere with filesharing software. Optimization of filesharing software, rather than an escalating battle, seems the sounder option for them.

How this conflict gets resolved is unclear. In the near term there is little that the ISPs can do about it. But at least they may start to become involved in the policy battles over copyrights and filesharing, and one could hope that the significant collateral damage inflicted by this fight, combined with its general futility, could lead to more intelligent policy. It is interesting to note that the adverse parties in this conflict (the big ISPs and major content providers) are the same parties that appear to be aligned on the same side of the Hulu-Boxee story I posted earlier. I don't see any interplay between the two issues at this point, but it's something to watch for.

As an aside, the article quotes the creators of the software as stating that its intent is, in part, to create a platform free from the prying eyes of an oppressive government. This is likely BS. While this software would make it considerably more difficult to track a file to its source, it is not impossible. With access to the PCs of the intermediate users (or, more likely, the logs of their ISPs) it is still possible to track a file back to its original source. While this is impractical in the context of a copyright claim, it is not something I would want to stake my life or liberty on where an oppressive regime is concerned.

ps. Public service announcement: This is unbelievably awesome.