Yes, except that businesses can't now make decisions that violate human rights codes. If they come up with mechanistic ways of doing that and it makes it impossible for us to know whether that's happening or not, they're still violating human rights codes. That was my point.
My point is that part of machine-based decision-making is that we don't get to see that, so we may miss discriminatory practices that we might otherwise pick up in human-based decision-making. One way of addressing that is the example from the GDPR that says, for certain kinds of decisions, you have to be able to provide an explanation. That will mean that you'll have to be able to provide an explanation that is humanly understandable.