This chapter describes algorithmic decision-making (ADM) systems. ADM systems are tools that leverage an algorithmic process to arrive at some form of decision such as a scoring, ranking, classification, or association that may then drive further system action and behavior. Such systems could be said to exhibit artificial intelligence (AI) insofar as they contribute to decision-making tasks that might normally be undertaken by humans. However, it is important to underscore that ADM systems must be understood as composites of nonhuman actors woven together with human actors such as designers, data-creators, maintainers, and operators into complex sociotechnical assemblages. If the end goal is accountability, then transparency must serve to help locate the various positions of human agency and responsibility in these large and complex sociotechnical assemblages. Ultimately, it is people who must be held accountable for the behavior of algorithmic systems. The chapter then highlights what is needed to realistically implement algorithmic transparency in terms of what is disclosed and how and to whom transparency information is disclosed.