An academic study found that large language models that drive some humanoid robots could make the machines prone to bias, ...