Natalie Gao, MJLST Staffer
The advent of the electronic age brought about digital changes and easier accessibility to more information but with this electronic age came certain electronic problems. One such problem is whether or not electronic entities like, (1) usernames online, (2) software agents, (3) avatars, (4) robots, and (5) artificial intelligence, are independent entities under law. A username for a website like eBay or for a forum, for all intents and purposes may well be just a pseudonym for the person behind the computer. But at what point does the electronic entity become an independent entity, and at what point does the electronic entity start have the rights and responsibilities of a legally independent entity?
In 2007, Plaintiff Marc Bragg brought suit against Defendants Linden Research Inc. (Linden), owner of the massive multiplayer online role playing game (MMORPG) Second Life, and its Chief Executive Officer. Second Life is a game with a telling title and it essentially allows its players to have a second life. It has a market for goods, extensive communications functions, and even a red-light district, and real universities have been given digital campuses in the game, where they have held lectures. Players of Second Life purchase items and land in-game with real money.
Plaintiff Bragg’s digital land was frozen in-game by moderators due to “suspicious” activity(s) and Plaintiff brought suit claiming he had property rights to the digital land. Bragg v. Linden Research, Inc., like its descendants including Evans v. Linden Research, Inc. (2011), have been settled out of court and therefore do not offer the legal precedents it could potentially have had regarding its unique fact pattern(s). And Second Life is also a very unique game because pre-2007, Linden had been promoting Second Life by announcing they recognize virtual property rights and that whatever users owned in-game would be belong to the user instead of to Linden. But can the users really own digital land? Would it be the users themselves owning the ditigal land or the avatars they make on the website, the ones living this “second life”, be the true owners? And at what point can avatars or any electronic entity even have rights and responsibilities?
An independent entity is not the same as a legal independent entity because an latter, beyond just existing independently, has rights and responsibilities pursuant to law. MMORPGs may use avatars to allow users to play games and avatars may be one step more independent than a username, but is that avatar an independent entity that can, for example, legally conduct commercial transactions? Or rather, is the avatar conducting a “transaction” in a leisure context? In Bragg v. Linden Research, Inc., the court touches on the issue of transactions but it rules only on civil procedure and contract law. And what about avatars existing now in some games that can play itself? Is “automatic” enough to make something an “independent entity”?
The concept of an independent electronic entity is discussed in length in Bridging the Accountability Gap: Rights for New Entities in the Information Society. Authors Koops, Hildebrandt, and Jaquet-Chiffelle compares the legal personhood of electronic artificial entities with animals, ships, trust funds, and organizations, arguing that giving legal personhood to basically all (or just “all”) currently existing electronic entities bring up problems such as needing representation with agency, lacking the “intent” required for certain crimes and/or areas of law, and likely needing to base some of their legal appeals in area of human/civil rights. The entities may be “actants” (in that they are capable of acting) but they are not always autonomous. A robot will need mens rea to assess responsibility, and none of the five listed entities do not have consciousness (which animals do have), let alone self-consciousness. The authors argue that none of the artificial entities fit the prima facies definition of a legal person and instead they moved to evaluate the entities on a continuum from automatic (acting) to autonomic (acting on its own), as well as the entity’s ability to contract and bear legal responsibility. And they come up with three possible solutions, one “Short Term”, one “Middle Term”, and one “Long Term”. The Short Term method, which seems to be the most legally feasible under today’s law, purposes creating a corporation (a legally independent entity) to create the electronic entity. This concept is reminiscent of theorist Gunther Teubner’s idea of a using a hybrid entity, one that combines an electronic agent(s) with a company with limited liability, instead of an individual entity to give something rights and responsibilities.
Inevitably, even though under the actual claims brought to the court, Bragg v. Linden Research, Inc. mostly seems more like an open-source licensing issue than an issue of electronic independent entity, Koops, Hildebrandt, and Jaquet-Chiffelle still tries to answer some questions that may be very salient one day. Programs can be probabilistic algorithms but no matter how unpredictable the program may be, their unpredictability is fixed in the algorithm. An artificial intelligence (AI), a program that grows and learns and create unpredictability on its own, may be a thing of science fiction and The Avengers, may one day be reality. And an AI does not have to be the AI of IRobot; it does not have to have a personality. At what point will we have to treat electronic entities as legally autonomic and hold it responsible for the things it has done? Will the future genius-programmer, who creates an AI to watch over the trusts in his/her care, be held accountable when that AI starts illegally funneling money out to the AmeriCorp bank account the AI was created to watch over, into the personal saving accounts of lamer non-MJLST law journals in the University of Minnesota? Koops, Hildebrandt, and Jaquet-Chiffelle argues yes, but it largely depends on the AI itself and the area of law.