In an attempt to ensure fairness and accuracy in facial recognition software across gender, age, and race, researchers at IBM suggest using "facial symmetry and skull shapes - to categorise faces" (O'Brien, 2019).

This proposal harkens back to craniometry, which is the 19th century "science" of measuring skulls to predetermine "morality" and "criminality," and its biological construction of "racial inferiority" based on skull size. Many proponents of craniometry argued that Black, Indigenous, Asian, and Latinx skulls are inherently smaller than European skulls because of their "inferiority."