Should Internet service providers block copyrighted material from their networks if “fingerprinting” technology allows them to easily identify it?
Bits, the New York Times’ technology blog, has an excellent copyright discussion featuring Columbia Law Professor Tim Wu and the general counsel of NBC Universal, Rick Cotton. One of the questions is whether ISPs should block copyrighted works? Wu makes a good point,
“Technologies designed to examine what kind of content is passing the network are technologies of censorship. Tolerating the routine inspection of all content, in the search for “forbidden” content, is a fast road to a private police state.”
But I think Wu is glossing over an important point here, i.e., there is a vast difference between a “private police state” and a real police state: Private entities are vassals of the state; the state answers to no one.
But, still, government can op-opt corporations to conduct surveillance for government.
Cotton is concerned about “wholesale reproduction of entire works (or even major portions)” and claims that: “as much as 50% of the carrying capacity of many, if not most, Internet service providers are being hijacked by a relatively small percentage of users to carry huge volumes of complete, unchanged pirated digital copies of films, games and software.”
There are a couple interesting questions here: One is whether common carriers have a right to block traffic? I think the answer is they can’t block lawful content, but audio and video files which violate copyright aren’t lawful.
Another interesting question is whether they can allow unlawful content if they have the technical means to block it? The Supreme Court has described a contributory infringer as one who “was in a position to control the use of copyrighted works by others and had authorized the use without permission from the copyright owner.” A good argument could be made that fingerprinting technology could block traffic which violates copyright. The question is whether the technology would also inadvertently block substantial lawful content?
Cotton believes the answer is no, but others disagree:
“The problem, according to [Electronic Frontier Foundation senior staff attorney Fred] von Lohmann, is not that fingerprinting isn’t good at identifying copyrighted content but that it identifies too much, including instances where a copyrighted work is legitimately incorporated into a new creative context as allowed under the doctrine of fair use.”
Speaking of the “Fair Use” Doctrine, Wu claims it is too vague and recommends a simpler formulation: “work that adds to the value of the original, as opposed to substituting for the original, is fair use.”
Wu’s suggestion would mean you could add a song to the video you shoot of your dog jumping into a bathtub, and the song wouldn’t be protected by copyright because it is part of your home movie. It’s one of those exceptions that swallows the rule… It would be great in terms of protecting Google’s YouTube from copyright liability, but are the rights of the secondary artist more important than the rights of the primary artist?
Why couldn’t Google, who earns billions of dollars per quarter in advertising, just pay the copyright-holder a reasonable fee when a YouTube video expropriates their song? It would be easy to tally the number of times the video is viewed and base the fee accordingly. Apple iTunes charges only 99 cents to buy a song, so YouTube wouldn’t have to pay much when one of it’s users merely listens to a song.
For more information on fingerprinting, see, e.g., “Policing Web Video With ‘Fingerprints’ – Sharing Sites Say Technology Could Help Them Identify, Remove Unauthorized Clips,” from the Wall Street Journal.