11 Pages Posted: 1 May 2007 Last revised: 27 Mar 2014
Date Written: April 27, 2007
A spate of papers have recently appeared raising the issue of the possibility of not only artificial agency, but also artificial moral agency that raises all sorts of substantive questions of moral responsibility. Suppose, for example, that we are epistemically justified in believing that an ICT engineer has designed and produced an ICT that is capable of acting and satisfies the criteria for moral agency. If this ICT turns out to do something bad that no one anticipated, who is morally responsible: the designer, the ICT or some combination of both?
What I wish to attempt to do is to work out some conceptual issues regarding the concepts of agency, natural agency, artificial agency, and moral agency, as well as articulate the criteria for moral agency as a first step towards beginning a consideration of these important questions of professional responsibility. Much of what I take myself to be doing enjoys a consensus in the literature - so much so that many crucial claims on which I rely are taken for granted in such widely-used and respected professional resources as the Stanford Encyclopedia of Philosophy, Routledge Encyclopedia of Philosophy, and the Internet Encyclopedia of Philosophy. I attempt to flesh out some of the implications of some of these well-settled theories with respect to the prerequisites that an ICT must satisfy in order to count as a moral agent accountable for its behavior.
I will begin with analyses of the more basic concepts, like that of agency, and work up to an analyses of the more complex concepts, like that of moral agency and rational free agency, subsequently considering the criteria something must satisfy to be accountable for its behavior; all of this will largely be uncontroversial in the literature. I will then argue that the each of the various elements of the necessary conditions for moral agency presupposes consciousness, i.e., the capacity for inner subjective experience like that of pain or, as Nagel puts it, the possession of an internal something-of-which-it is-is-to-be-like. I ultimately conclude that the issue of whether artificial moral agency is possible depends on the issue of whether it is possible for ICTs to be conscious.
Keywords: action, agency, artificial agency, moral agency, moral accountability, consciousness
Suggested Citation: Suggested Citation
Himma, Kenneth Einar, Artificial Agency, Consciousness, and the Criteria for Moral Agency: What Properties Must an Artificial Agent Have to Be a Moral Agent? (April 27, 2007). Ethics and Information Technology, Vol. 11, No. 1, 2009. Available at SSRN: https://ssrn.com/abstract=983503 or http://dx.doi.org/10.2139/ssrn.983503