Computer ethicists have long been intrigued by the possibility that computers, computer programs, and robots might develop to a point at which they could be considered moral agents. In such a future, computers might be considered responsible for good and evil deeds and people might even have moral qualms about disabling them. Generally, those who entertain this scenario seem to presume that the moral agency of computers can only be established by showing that computers have moral personhood and this, in turn, can only be the case if computers have attributes comparable to human intelligence, rationality, or consciousness. In this chapter, we want to redirect the discussion about agency by offering an alternative model for thinking about the moral agency of computers. We argue that human surrogate agency is a good model for understanding the moral agency of computers. Human surrogate agency is a form of agency in which individuals act as agents of others. Such agents take on a special kind of role morality when they are employed as surrogates. We will examine the structural parallels between human surrogate agents and computer systems to reveal the moral agency of computers.

Our comparison of human surrogate agents and computers is part of a larger project, a major thrust of which is to show that technological artifacts have a kind of intentionality, regardless of whether they are intelligent or conscious. By this we mean that technological artifacts are directed at the world of human capabilities and behaviors.