Wednesday, July 22, 2009

Creative Commons and the Curse of Verification

I am a huge fan of Creative Commons. Many people do not know that I was actually the one that initiated the development of the Australian Creative Commons Licenses. I attended the internet law program at Stanford University in 2003 where Lawrence Lessig lectured on internet regulation and the Creative Commons project. I told Professor Brian Fitzgerald at the Queensland University of Technology about it and he agreed to supervise my independent research project the next semester on this area. I mention this as a prelude to what I am writing about today because I want to be clear that I support this organisation and am very much in agreement with their goals and ethics (note that both of my blogs are licensed under Creative Commons Licenses).

It is with some concern that I have read that the new licence verification system introduced to Google Image Search which is designed to assist users to find and use images which are openly licensed, is proving to be less than successful. Many people have licensed works incorrectly and as such the technology is reported to be producing a number of false positives – that is - listing files that are in fact under copyright.

A few years ago, over the course of a year (2006), I attempted a statistical analysis of Creative Commons Licenses in Limewire and much the same took place. The Limewire software, in co-operation with Creative Commons, enables people to tag files as being openly licensed and allows users to verify the existence of these licenses. Right clicking on a file gives the user the option of ‘viewing the file’ and when viewing the file the user has the option of 'verifying the file'. By clicking on the verification link the software automatically goes to the web page that the uploader has given and checks that the license is on the given page.

More times that I care to state this technology was faulty. Many times uploaders incorrectly tagged files as being openly licensed when they were not. Other times the verification link would link to a page that had a Creative Commons license but was not the correct page for that file. These results, while highly disruptive and very very annoying, can be attributed to uploaders being ignorant or deliberaltely doing the wrong thing. However in other circumstances, the verification link would report that the file could not be verified and yet when one went to the web page that was listed with the file, there was in fact a Creative Commons license on the page and it was for the correct artist.

Reading today about these similar problems with image verification brought back the intense frustration I felt at the time of doing that statistical analysis. What I was trying to establish (with, as it turned out, less than perfect research design) was that there was a growing trend of correctly licensed Creative Commons files. My results did in fact show an increasing trend, however the process or method involved in verifying each license was exhausting and time consuming because I could not rely at all on what the technology told me. Before writing this blog post I went into Limewire to see if it had changed but it hasn’t. There were many files which were clearly uploaded and tagged incorrectly and others that did not verify that should have – Brad Sucks is on Jamendo and is openly licensed but the verification process in Limewire will tell you that it can't be verified.

Many a time I sat there and questioned the effectiveness of this technology.

And yet here again a similar mechanism has been put in place and the system is reported to be achieving the same results.

My concern here is two fold.

Firstly I worry about the artists that are genuinely trying to use open licensing and file sharing networks as part of their business or communication model. This mechanism does little to make the process of negotiating a file sharing network like Limewire much easier. While it is good to have the search mechanism in place the results are far too inconsistent to suggest that it is sufficiently effective.

Secondly I worry about Creative Commons. From my view these imperfect technologies open the organisation up to a secondary copyright liability claim. Just as file sharing networks state that it is up to the public to use their technology lawfully and that they are not responsible for the illegality that takes place, so too Creative Commons appear to distance themselves from responsibility. However there is scope in my mind to suggest that knowingly producing and promoting technology that has the potential to ‘induce’ or mislead a user into thinking something is licensed when it is not, is a dangerous game to play.

Just as the likes of Lawrence Lessig (and myself) advocate for the registration of copyrights, so too should Creative Commons be doing more to ensure that works are correctly licensed. While I understand the logistics of this are immense and the costs prohibitive, the answer does not lie in releasing technology that works, at best, once in a while. I appreciate the need for these search mechanisms and they are in some ways better than nothing but the risk here is that the whole organisation could be held liable for secondary copyright infringement. It would be an unpopular lawsuit but the disclaimers are simply not enough.

In the USA Grokster case and in the Kazaa case in Australia the same arguments were made – that the users were responsible, that it was not the primary purpose of the technology, and that they disclaimed liability; and this was not enough. I hope that those at Creative Commons have better arguments than these otherwise there could be trouble ahead.

I am a huge supporter of this organisation please don’t mistake this criticism as anything other than genuine concern. This litigious environment has had many casualties and I would hate to see this organisation on that list.

Further Reading
The Register, The tragedy of the Creative Commons (16 July 2009) <> at 22 July 2009

No comments: