Scholars have increasingly focused on understanding different aspects of algorithms since they not only affect individual choices and decisions but also influence and shape societal structures. We can broadly categorize scholarly work on algorithms along the dimensions of economic gain that one achieves through automation and the ethical concerns that stem from such automation. However, the literature largely uses the notion of algorithms in a generic way and overlooks different algorithms’ specificity and the type of tasks that they perform. Drawing on a typology of tasks based on task complexity, we suggest that variations in the complexity of tasks contribute to differences in 1) their automation potential and 2) the opacity that results from their automation. We also suggest a framework to assess the likelihood that fairness concerns will emanate from automation of tasks with varying complexity. In this framework, we also recommend affordances for addressing fairness concerns that one may design into systems that automate different types of tasks.