Algorithm bias refers to the systematic and unfair discrimination that is built into algorithms. This can occur when data used to train AI models is incomplete or reflects existing biases. Understanding and mitigating these biases is critical for creating fair technology.