Corporate Moral Agency and Artificial Intelligence

1. Introduction

For about half a century a debate has taken place about the application of moral-philosophical categories to corporations and collectivities. It has been wide-ranging self-critical and inconclusive. Within academic business ethics the debate has focused upon the very idea of the corporation as a moral agent (CMA) and the implied responsibilities and rights of business corporations per se. Contributions have come from several academic disciplines, including legal philosophy and politics (where it is sometimes framed as a debate about corporate citizenship) but also the cognitive and systems sciences.

More recently, a somewhat similar debate has been taking place about the very idea of artificial moral agency (AMA) or “moral machines” (e.g. Wallach & Allen 2009). The intention in this case is to clarify the potential moral rights and responsibilities of artificial general intelligence (AGI) systems per se but also the individual (humans) who contribute to their construction. The present paper accordingly considers the overall relationship between the CMA debate (CMAD) and the concept of AMA.

In carrying out this exercise it is worth bearing in mind from the outset that the CMAD can be directly interpreted as a debate about artificial moral agency (AMA) when one considers the hypothetical case of a virtual firm: one that is being run by AGI’s (Figure 1). For the general case, however, the paper proceeds by suggesting various ways in which arguments within (and about) the CMAD might inform the very idea of AMA or else guide ethical aspects of the AGI-building (AGI/B) project¹.

Figure 1.

The CMA debate can inform the idea of AMA

In the following section, the CMAD in business ethics is briefly reviewed and a way of structuring it is suggested². Various arguments for or against CMA are duly distinguished from meta-arguments that critique that entire debate itself. Sections 3, 4 & 5 then set out the core of the paper in which all such arguments are related to AMA. The paper concludes with a brief discussion of philosophical pragmatism and the prospect of artificial-ethics.

2. Structure Of The Cma Debate

The corporate moral agency debate (CMAD) is comprised of various pro-CMA arguments that support and endorse the very idea that corporations are moral agents (pro-CMA) whilst other (anti-CMA) arguments critique or qualify it. Several other contributions are in the form of meta-arguments that qualify the importance of the entire CMAD (Figure 2).

Figure 2.

The corporate and artificial moral agency debates

Within the CMAD, the pro-CMA and anti-CMA arguments can be classified in various ways. Firstly a distinction can be drawn between:

• 1.

  legalistic and political arguments that focus mainly upon the distribution of responsibilities duties and rights; and

• 2.

  psychological and systems-theoretic arguments that tend to dwell on individual and corporate capabilities and their implications.

Arguments in each class then appear to carry distinctive implications for the concept of AMA and for the AGI project. In particular, the “systemic and psychological” arguments that involving cognitive limitations also seem to imply that AGIs (or AGI-run corporations) have the potential to be morally superior to human-run corporations (Figure 3).

Figure 3.

A typology of arguments