And AI models just go with their programming. Like humans they may think they're making unbiased decisions but a lot of work has went into aligning them to give specific kinds of answers. We're all subject to biases based on what we were taught and how we want to be perceived.