Artificial intelligence (AI) systems have the potential to perpetuate racism and bias, according to experts. Instances of AI displaying racist tendencies have been observed, such as a tool that generated images of Barbie dolls, resulting in a German Barbie wearing a Nazi SS uniform and a South Sudanese Barbie holding a gun. Nikon’s face detection software in 2009 would also ask Asian people if they were blinking. In 2016, an AI tool used by US courts predicted twice as many false positives for black defendants as white defendants. The UK data watchdog, Information Commissioner’s Office (ICO), is investigating the issue, stating that it could have damaging consequences. However, experts argue that there is nothing inherently racist about AI and that the issue lies in the data sets used to train the AI models.