I'm not convinced that's reasonable, really. If you say "I hate niggers, they ruin everything", that's easy. If you say "Hilary Clinton's aides are raping kids in a pizza restaurant", that's easy too. But these employees checking it can't really just be random goons. If I post an article that I found (or maybe that I wrote on my blog) about how I think the President of South Korea's father is embezzling money through private finance initiatives, it could be total lies. It could have a kernel of truth surrounded by lies. It could have a kernel of truth surrounded by lots of conjecture. It could be totally true. It could be a next weeks big news story on all the major networks, or it could be a total work of fiction. I don't think an algorithm could realistically determine this, and even if it flagged it, who's going to determine that? A Facebook employee?