[delayed]

You are right and I am speaking as someone who has worked on these systems. It's hard to convey to people, as these systems scale up, the bugs scale up too. And the resources to handle them shrink. Even if the false positives are less than 1% tens of thousands of people can get effected. And things are swept under the carpet, cause the costs to corporations are rounding errors.

> I don’t think taking the banks answer to any of those questions at face value is responsible citizenship

I agree but I wanted to push back on the idea that this was somehow due to automated decisions being made. That isn't really the problem here. The same exact thing would happen if the government mandated "no computers are allowed to be involved". The problem is opaque decision making, no clear escalation path, a disregard for the short term pains, and so on.

The fundamental problem isn't how the decision gets made. Even if a human made the decision you'd have the same issues of the front line consumer support saying "well the other department made their decision, I don't know why and I don't have the authority to change it and they don't accept calls".

By talking about automated decisions, the OP is making the conversation about the wrong problem.

I can agree with this. I do think automated decision making amplifies the effect, but you are right on the fundamental issue