It's something that can be overcome with role-play training and consistent reinforcement. Examples are actually given in the book. One case study is a Korean ai

It's something that can be overcome with role-play training and consistent reinforcement. Examples are actually given in the book. One case study is a Korean airline that had a horrifying safety record in part because copilots and other staff didn't feel empowered to overrule the captain even if the captain was wrong. It was a cultural problem, but was overcome through training.