"The Pope endorses Trump" would be labeled as fiction. But what about "The Pope endorses [some of] Trump's views"? Or "The Pope is a bigger fan of Trump than you think"?
In fact, they do the opposite: they promote it to millions of people. Or more accurately, they permit their software to do so. And they claim they have no responsibility, even though they built the entire thing.
I saw the BS News extension it checks each link for citing known fake news sites in their article.
It is not practical as I keep seeing new fake news sites popping up from nowhere. Just register a domain, install Wordpress, write some articles and then pay for advertising on Facebook.
Fake news sites got Trump elected by saying false things like the Pope supports him. It is serious when people only use Facebook as their only news site and read and believe without checking the facts first.
If I were to make a fake news site detector I'd have it search Snopes and Politifact and others to see if it has been already reported as false.
Basically the way it ends up is a new way to censor the news using AI.
Suppose instead of a black list of sites they use a white list of trusted sites? Then block any new sites until they can verify them as trusted?
Decisions with political implications are not the same thing as political decisions. Huge categorical difference, in fact.
That has political implications because everything does. But evaluating content based on those implications, rather than on the actual facts, is a categorical error.
We label fiction and nonfiction in libraries, book stores, movies, magazines, etc--basically every other medium for accessing content. Why not Facebook too?