In their very interesting draft paper that Dan mentioned, Professors Bradley and Froomkin suggest that virtual worlds might be testbeds for legal rules. However, with regard to experimenting with intellectual property, they aren't so sure that virtual worlds are a good fit. They say:
Like zoning, intellectual property involves complex technical issues in the real world that a game could not test.
This may be true, but a bigger problem with using virtual worlds as testbeds for experimental intellectual property rules is that virtual worlds are intellectual property. Putting aside trademarks, patents, and other relevant forms of intellectual property, software is protected by copyright. The copyright is not just limited to a game's source code and object code, but also extends (to an unclear extent) to other salient aspects of the program.
With regard to intellectual property rights in avatars and MMORPGs, Ren Reynolds, Molly Stephens, and Dan Miller have all looked at the issue. And Joseph Beard has a funny and well-written piece that isn't directly on point, but is worth reading to understand the greater context of virtual property rights. I'm not going to foray into the legal issues here in a blog post -- I'll just say that IP issues in virtual worlds are very complicated and, as a practical matter, much depends on the language of click-wrap licenses and the degree to which those licenses might be enforced.
What I did want to say is that we should bear in mind how intellectual property issues and governance issues in game spaces overlap. T.L. Taylor has a very interesting article re ownership in Everquest, that demonstrates how governance and IP ownership issues can be intertwined. As Taylor and others have noted, it can be argued that the participants in MMORPGs are like actors who give life to the world. But even if that is true, huge investments of creativity (as well as money) are made by the designers (and corporations) who set the MMORPG stages for the actors, and then market, distribute, maintain and improve them.
Dan and I pointed out in our article that analogizing virtual worlds to real life legal regimes is problematic because we don't usually need to credit our elected officials with the authorship of our worlds, whereas we usually accept the fact that authors and artists "rule" absolutely the worlds they create. (Admittedly, some fan feedback websites to the contrary.) Therefore, seeing games as akin to governments, while it may be appealing at first glance (especially to players) is hard to reconcile with the rights of authors. Pavel Curtis suggested that in virtual worlds, dictatorship is inevitable. It will be interesting to see, therefore, to what degree democracy, authorship, and intellectual property can reach a market equililbrium as these environments continue to develop.