So this actually sounds like a realistic approach at managing the dangers of AI. It ensures that people have some sort of recourse against algorithms deciding t

So this actually sounds like a realistic approach at managing the dangers of AI. It ensures that people have some sort of recourse against algorithms deciding to make their life miserable. This feels like an extension of similar approaches used in credit: credit rating agencies have to allow you to look at the data used, and are required to have flows for people to challenge the data there that may be harming them. Certainly it's a very different approach from people trying to mandate that AIs must be designed in ways so that they can't be used for bad stuff (which to me feels like a fundamentally broken approach).