Is it surprising? Google and Facebook can barely police the content their human users post.

That seems to be by choice, not because it's inherently difficult at scale. It's not like we see them take things down when you report them, you'll get multiple reviews of "this obvious scam doesn't violate our rules".