Right. And the argument is that the minute Facebook actually reads your post and makes a decision about how much exposure it will get based on its content then it is Facebook posting things.
You could make the same argument about Google search, which is a decidedly more user-driven algorithm determining what you're seeing.
Google can't and shouldn't be responsible for the factual accuracy of sites that it indexes and shows to users.
Consider e.g. what happens when libelous content is indexed.
AI generated content is a different story, and I would not expect a court to lean towards Section 230 protections for such content. But user generated content, or content index, scraped, or otherwise reproduced verbatim from real humans is a different story.
If you want publishers penalized for revenue generation that comes from illegal content, then it seems appropriate to draft a law that requires attribution of said revenue to specific pieces of content, and the ability to use this in discovery.
That thinking strikes me as "clear, simple, and wrong". (I appreciate that 230's bright-line rule may also be clear, simple and wrong)
Consider defamation. Often, the difference between a defamatory and non-defamatory statement is truth. Expecting websites to distinguish true statements from false ones is a non-starter.
Let's say my family has a horrible experience with a youth pastor. I post about it on facebook to warn people in my community. If my claims are false, they're almost certainly actionable defamation. If my claims are true, disallowing them to mitigate Facebook's potential liability is also bad, but not in a way that affects Facebook.
