Banning spam might be possible without giving platforms the power to make their own judgements about the truthfulness or decency of the content they host.
If 90% of (a random subset of) users agree that a given piece of content is spam, the platform should be entitled to delete the content. The company would then be allowed to ban a user after a certain number of strikes, possibly subject to an appeals process where a human employee checks that this isn't a case of a minority viewpoint being unfairly silenced by false reports.
This would democratise these platforms, and only give companies the discretion to allow more content than their users are interested in, rather than less.
