Airbnb uses cyberstalking to go through people's social media accounts to see if they're psychopaths. While the goal of safety is laudable, stalking people all through the internet is the opposite of safe. And it's not just one business doing this -- it's more and more common. This contributes to the overall anxiety of people who are constantly being examined and criticized, which is ruinous to health. And does the algorithm only exclude people who are objectively dangerous (e.g. criminals)? Or does it also exclude all the unpopular folks like black people and autistics and poor people and so forth? AI tends to absorb the bigotry of its programmers, so the outlook is not good.