Volume 2, Issue 1
1st Quarter, 2007


Artificial Intelligence as a Legal Person

David Calverley, Esq.

This article was adapted from a lecture given by David Calverley, Esq. at the 2nd Annual Colloquium on the Law of Transbeman Persons, on December 10th, 2006 at the Space Coast Retreat of Terasem Movement, Inc., Melbourne Beach, FL.

 As a private practice attorney and biotechnology enthusiast, David skillfully explores the frontier of complex legal, philosophical, ethical, and ‘personhood’ issues society universally faces in the coming, technological  age of Artificial Intelligence.

I'm often accused of taking on intractable subjects and trying to deal with them and then finding out that I cannot do justice to all of the nuances. I think you've seen the risk of that today because the factors that we've talked about implicate political theory, moral philosophy, and law. We could go on for months, if not years, about some of these issues.

In addition to thanking Terasem for allowing us the opportunity to begin these kinds of discussions, I'd also like to thank “AI and Society” Magazine. The paper that this talk is based on will be published by “AI and Society” in a special issue this spring. Thank you to them for permission to draw heavily on that paper.

Also, I'll probably go into some issues that are in an article that has just been published in Connection Science called, Animal Rights, Does An Analogy Exist?

I'd like to start with some background information, and I'd like to look at and ask the question:  What is law? What I'm going to try to do is discuss this topic more from a universal perspective rather than limiting it to, for example, the U.S. Constitution, not for any particular reason other than I think that the U.S. Constitutional definition of person perhaps has some nuances that may not apply in other legal systems.

 


Image 1 - Legal Person

What I'm really trying to tease out is what are the things that we call law and who does law affect. I want to start with a definition from Steven Morse drawn from an article[1] “New Neuroscience, Old Problems” which appears in a book Called “Neuroscience and the Law”. "Law is a Socially constructed, intensely practical evaluative system of rules and institutions that guides and governs human action, that help us live together. It tells citizens what they may, must and may not do, and what they are entitled to and it includes institutions to ensure that law is made and enforced."

”Why do we care? As researchers in artificial intelligence or in transbeman processes we care because that is the kind of definition that society will apply to us.  It will look at us as something on which the law can act.

If one of the goals that we have as AI designers is to create something that comes into the category of human equivalence, then we need to look at the process that other people will analyze us by.  As Larry Solum[2] indicated, "Putting the AI debate in a concrete legal context acts as a pragmatic Occam's razor[3].  We are forced to see anew in a relentlessly pragmatic context what is it that we mean when we say that artificial intelligence has legal rights."  


Image 3 - Why do we care

From that I think we can make some arguments in a broad theoretical sense and try to really understand what we mean as human beings when we talk about a legal person.  To do that I am drawing on Folk Psychology as a mechanism to allow us to analyze some of these ideas that we are talking about.

 
Image 4 - Folk Psychology 

Let me back up and state that in terms of Folk Psychology what we are really looking at is what you and I and other people generally believe when we act. Law is, if nothing else, intensely practical, it assumes that people are practical; reasoners, and it is based on a Folk Psychology model. It recognizes and argues that people act in intentional ways. They create desires, they have beliefs and they act on those.

Where law comes into the picture is with respect to a determination that someone has formed a belief, formed a desire, and acted upon that. If that action is an inappropriate action, then the law will punish that action.

What we need to look at then is: what is this intention? The whole process is in fact, hierarchical. After the formation of a desire you have to have the capability to perform that desire. You have to have underlying beliefs from which those desires draw their motive power, but it can't stop there. It has to become action driven.

Intentionality, in philosophy, has a slightly different meaning, and I'm hopefully not confusing the two. Intentionality for a philosopher simply means the aboutness of something.  If I make the statement that Cape Canaveral is in Florida, that's a statement about Cape Canaveral.  It's also a statement about Florida and the situation of Florida actually containing this thing that we call Cape Canaveral.  


Image 5 - Intentionality 

Intentionality for Folk Psychology purposes, is slightly different, so we need to keep the two separate. Intentionality in this sense is the belief, desire, and act process. And then there is one additional thing I did not mention, there has to be the skill to perform and the awareness that a given action is being performed.  So if someone wanted to fly to the moon and formed a belief and a desire and jumped off a bridge, they don't have the skill. So it would not fall within that sense of intentionality that really is encompassed in that definition.

How does that apply to our discussion? What we are dealing with is a situation where if we can argue that a nonbiological machine can form an intention derived from a desire, and a belief and can act intentionally based upon that intention, at least from a theoretical standpoint, there seems to be no reason why that entity could not be viewed as a legal person.

That argument has been carried through to corporations, Peter French, ‘writing in the late' 80s, developed a long argument that corporations, in fact, act intentionally in a Folk sense, and that they do create action based upon belief and desire, obviously have the skill to perform that and the awareness that they are, in fact, doing that.[4]

So there is at least from that perspective, an argument that nonhuman, non-homosapien entities can in fact, form these kinds of intentions. Again, referring to a quote from Larry Solum, “How would the legal system deal with the objection that the AI does not really have intentionality despite its seemingly intentional behaviors?  The case against real intentionality could begin with the observation that behaving as if you know something is not the same as really knowing it."

Next Page

Footnotes

[1] Morse, S. (2004). New Neuroscience, Old Problems. Neuroscience and the Law. New York: Dana Press.

[2] Solum, L. (1992). Legal personhood for artificial intelligences. 70 North Carolina Law Review. 1231.

[3] Occam’s razor – paraphrased, "All things being equal, the simplest solution tends to be the best one." http://en.wikipedia.org/wiki/Occam's_Razor  February 20, 2007 12:12 PM EST 

[4] French, P. (18840. Collective and coporate Responsibility. New York: Columbia University Press

1 2 3 4 next page>