The role of AI systems in the workplace is growing. Many tasks will be delegated to such automated decision-making systems that rely on data as a training basis of knowledge. Data comes with biases that can remain undiscovered during the lifecycle. AI systems need to be designed in a way that considers the incompleteness of the information they rely on. Name, define, measure, and identify the effect of bias on fairness, which are upcoming tasks for regulators to support AI developers in managing risks. Biases might affect societal norms, fairness, ethical assumptions, and decision-making ultimately. The still limited methods for mitigating distortions are becoming available at an ever-faster pace. In this review article, we want to analyze the relevant literature and the central concepts of bias. After an introduction, we are developing a common understanding of what bias is. We outline the implications of bias for the workplace and how regulators in Europe and the US address AI and tackle the challenges regarding fairness. The ethical considerations underlying the regulatory activity and the public awareness of bias complete this review. Future directions and conclusive remarks are intended to help understand the path that AI regulation has taken for work systems.