> This is why we established centralized AI ethics governance through an AI ethics board at IBM.<p>Later on in the article it starts talking about legislation.<p>To me this is about A: building some bias into the model (which people either don’t perceive as bias because it’s their opinion or because they know it’s more PR friendly), and B: competing not by being the best but by being big enough to self hinder with these arbitrary rules and then lobby for legislation.