Abstract

This paper considers the proposed development and application of specific ethical principles to Artificial Intelligence (AI). It contrasts the approach with the application and evolution of orthodox public policy founded on welfare economics and markets, which also has ethical underpinnings. If new ethical principles are thought appropriate, they should be assessed against the counterfactual of existing public policy. Further, the appropriate focus for the application of any new principles should be assessed, are they specific to all AI, a sub-set of applications or technology and markets including human decisions more generally? It is far from clear that a general-purpose technology such as AI per seshould be the focus of new regulation. It is concluded that whilst changes are required – to remove barriers to AI, reconcile existing law and regulation with technology change and to adapt the rule making process to fast changing markets – it is unclear what purpose specific ethics for AI would serve.