I was be inclined think the same, until I remember how Google went on stage with their AI image generator that was banned from generating white people, to which I thought there was no way someone didn't notice such a bias programmed in before launch, or how Alexa would refuse to tell you good things about Trump when asked, claiming it doesn't discuss politics, but would gladly tell you all the good things about Kamala.

The most likely explanation is that stuff like this was indeed noticed by other people but who shared the same ideological and political biases and wanted this to go into production as a way to "own the chuds", or they knew it was wrong but dared not to speak up for fear of being labeled "-ists" or "-fobes" by the ideologically driven mobs in the org.

It's not just one rogue individual with a bias getting away with it, since there's too many points on the graph already, and they're all forming a line pointing in the same direction.