Algorithms are data-driven formulations. The Garbage-in, Garbage-out (GIGO) principle means that results from algorithms can only be as reliable as the data on which they are based. If there are biases in the data, they will be reflected in the algorithms as well.