Ethics and AI Issues: Old Container with New Wine?

This paper reflects on what differentiates AI ethics issues from concerns raised by all IS applications. AI ethics issues can be viewed in three distinct categories. One can view AI as another IS application like any other. We examine this category of AI applications focusing primarily on Mason's (MIS Quarterly, 10, 5–12, 1986) PAPA framework as a way to position AI ethics within the IS domain. One can also view AI as adding a generative capacity producing outputs that cannot be pre-determined from inputs and code. We examine this by adding "inference" to the informational pyramid and exploring its implications. AI can also be viewed as a basis for reexamining questions of the nature of mental phenomena such as reasoning and imagination. At this time, AI-based systems seem far from replicating or replacing human capabilities. However, if/when such abilities emerge as computing machinery continues growing in capacity and capability, it will be helpful to have anticipated arising ethical issues and developed plans for avoiding, detecting, and resolving them to the extent possible. [ABSTRACT FROM AUTHOR]