> when everyone is trying to put their best foot forward

Except they're not. A significant fraction of applicants are people you would not want in your company. Outright frauds. You find out when you are on the hiring end and you can see the raw applications without any filters. The question is are you going to reject them based on whatever information you can glean without a call or interview, or are you going to give them a chance? A looser screen is more democratic, but it calls for scalable solutions like this. Perhaps a middle ground is to screen only the suspect candidates with AI.