Welcome to a brave new world of digital fascism and mass surveillance. Why does every message have to be scanned with AI? Since when WeChat behaviour has become an example for the Western social media? You don't scan all user private messages with AI, you give your users a report button which they can press if they don't like a message. Only then you can take any other measures. That's is.
To answer this question honestly - I would assume that the service owners don't want to get pulled in front of a congressional inquiry about why their service isn't doing anything to prevent the distribution of child abuse material.