Coverage bias is a form of selection bias that occurs when the data used for analysis or training does not adequately represent the population or situations of interest. In AI systems, it can lead to poorer performance or unfair outcomes for underrepresented groups.
See selection bias.