Protecting against malicious use of DCMA notices
[Note, this post was prepared on Oct 15th but for reasons that will become obvious, posting was delayed until today]
Whatever else may be said about the 2008 US Presidential election, the Web did the job we intended it to do back in 1992. The 2008 election was not the first election in which the Internet (and indeed the Web) were used. In fact I ran a Web server with material from all of the parties back in 1992 and the Clinton-Gore campaign had an online campaign in that election run by Jock Gill.
But 2008 is the first election in which the agenda was not entirely set by the establishment media which would much rather debate lipstick on a pig than healthcare, education or the economy.
So much for self congratulation. Now to look at what did not work.
One of the biggest problems was that at the same time the Web makes it easy to make information available, it also makes that information less easy to trust. Unlike the 2004 election the 2008 election has not had whole media cycles dominated by fake photographs (Kerry/Fonda) or forged documents (aka Rathergate). But that seems to have been as much because people are much less likely to trust the information they see.
The good part is that we are less likely to be fooled. The bad is that we are less likely to be informed. We need to have a mechanism that allows people to actually trust their eyes.
This problem has many aspects and will take many years to solve completely. But a good starting point would be to look at the issue of copyright and in particular the use of DMCA takedown notices to supress speech as happend during the campaign.
The basic facts are that the McCain/Palin campaign uploaded videos to YouTube. YouTube then received DMCA copyright notices and the videos were removed. Whereupon the McCain/Palin campaign complained that their free speech was being suppressed and that videos from the campaigns should be vetted manually. In response to which Google said that they were not going to create special categories of content.
It is very easy to get into arguments about which side is right and overlook the fact that these goals are not necessarily incompatible.
In particular, the real problem here is a failure of accountability. The DMCA was written to address the problem of copyright infringement by unknown, unaccountable parties. It contains a provision for an objection to be made against a takedown notice. The whole point of DMCA is to identify the parties to a dispute so that it may be resolved in a court of law.
So why not allow any party to sign their uploaded content with a digital certificate that contains their authenticated business address? We have all the infrastructure in place to issue such certificates today (although Google can hardly be expected to have the code to make use of them.)
Comments
One reason that abusive takedown notices are such a significant problem is that if a service provider fails to act on a DMCA takedown notice (for example because they believe it is invalid) or restores content without a counter-notification, they are accepting liability for infringement. Because of this, no company acting within the jurisdiction of the United States will vet takedown notices. A joint USC/Berkely report in 2005 found that over 30% of DMCA takedown notices were illegal (the majority of those cases were legal fair use), and that nearly 60% were sent by a business targeting a competitor. Rather than trying to help content providers identify uploaders, it seems it would be a better approach to dissuade them from "crying wolf" by making companies accountable for the notices that they send, including significant fines for improper takedowns and financial liability for any losses suffered by an uploader as a result of the time their content was unavailable or other costs from the takedown request.
Posted by: Hunter Schwisow | November 10, 2008 12:44 PM