[bdw5204]

That's why there should be laws requiring large services to provide due process for users who have invested significant time and energy into the platform.

Its just common sense that a spam bot should be nuked on sight but a legitimate user should get a fair hearing before you ban them. Ideally the social media platforms would do something akin to Wikipedia's arbitration committee (basically the Supreme Court of Wikipedia) but scalable by hiring a suitably diverse pool of "unskilled" people for a $15 an hour or so remote job where they decide if a longtime legitimate user should be banned. I think this system would do a much better job of avoiding false positives. A properly functioning spam bot detector shouldn't be detecting any legitimate longtime users whatsoever.

Banning email addresses without sufficient notice to allow accounts tied to that email to be moved to another email should be illegal. You should also have a legal right to access your digital purchases even if your account is banned. There's nothing wrong with not wanting a customer any longer but you shouldn't be able to cause irreparable harm to your customer on their way out the door and there should be safeguards in place to ensure you don't fire any customers by mistake.