It is but it takes the blame off of the platform entirely.

To word it a different way - if I made a service where no matter what you wanted to say I would write down your words, read them, and run through town shouting them until I found the person it was directed to and then shout it at them… I would be liable for the words being said as much as the person paying me to say it.

If, however, I have a strict policy where I will only do the above after I strictly review and moderate your words, and you managed to sneak in a tongue twister that says something dirty that I didn’t realize until after I shouted it… I am no longer liable. I did everything a reasonable person could expect, you are the only one liable.

When people sue in the US (and when companies really fuck up) people sue the person liable and all possible parties that could be included. The parties then shift blame around pretrial and try to prove they are not liable by xyz to get dismissed off the case. If this fails then each party sued essentially has a trial for their specific liability, which needs to be separately proved in court; and if it makes it that far, in front of a jury (or panel of judges, or a single judge, depending on which state and what kind of action).