Ethics, Regulation and the New Artificial Intelligence, Part I: Accountability and Power

A generation ago, there was a major debate about the social and ethical implications of artificial intelligence (AI). Interest in that debate waned from the late 1980s. However, both patterns of public risk perception and new technological developments suggest that it is time to re-open that debate. The important issues about AI arise in connection with the prospect of robotic and digital agent systems taking socially significant decisions autonomously. Now that this is possible, the key concerns are now about which decisions should be and which should not be delegated to machines, issues of regulation in the broad sense covering everything from consumer information through codes of professional ethics for designers to statutory controls, issues of design responsibility and problems of liability.