Lots of clever people unintentionally trained deep networks that unfairly discriminated against people of various races or from certain zip codes. Sometimes you aren't aware of all the implications of the decisions you might make and a course that informs you of such pitfalls can be helpful. Professional ethics courses at the University level aren't there to convince you discrimination based on gender, race etc is wrong, but to make you aware of when that may be occurring.
I don't know. From what I remember of the flame war the problem was that they fairly discriminated against various races and zip codes. So actually the algorithms had to be made less accurate to fit the legal expectations.
But once again the problem was in the customers. They should have said optimize for a, while keeping discrimination in check, not just optimize for a
