Lack of data on race and ethnicity leads to biased AI.
In the context of race and racism, data helps validate experiences and perspectives that are often undercut, misunderstood, silenced, and ignored. It took until 2020 for Facebook to dedicate internal teams to study its main social network and Instagram for racial bias, and in particular, whether its algorithms trained using artificial intelligence adversely affect Black, Latine, and other underrepresented groups.