• Software will use artificial intelligence to accurately identify prohibited speech and automatically censor online posts. This is what they are doing now but suffers from false positives, censoring speech that should not be censored in a world that believes in free speech. The general vibe is that in a world of social media, false positives – and censoring valid speech – is acceptable.
• Require posters to register with a government or officially recognized authority in order to log in. Some nations are doing this now.
• Facebook has proposed that purchases of political ads on Facebook must be confirmed via a mailing to a U.S. postal address. This is so easy to get around it is silly.
• Require posters to be licensed (slightly different than registration).
• Charge a small fee for each post or share. This would discourage the viralness of propaganda memes, except those from billionaires. (My spell checker suggest “evilness” in place of “viralness”).
• Let users of social media flag content for review. This is already done and is actively used to censor political speech with viewpoints others disagree with. In other words, this doesn’t work but it is being done anyway. This is censorship via proxy – a group can gang up on someone’s posts and then get a 3rd party – the social media company – to censor the speech.
• Related to the above, is simply ganging up on individuals. Anyone who posts content with “wrong public opinion” is ganged up on, directly, by other posters and thoroughly ostracized. Or worse, doxxed by filing false police reports leading to police raids on innocent parties. This is already done routinely, worldwide.
• Combinations of the above are used and will be used to identify “fake news”. Fake news, of course, is different than valid political speech. Such algorithms will have a high false positive rate, flagging legitimate posts as fake news.

The best and most effective propaganda is based on facts. It works by leveraging the targets pre-conceived biases and by providing selective information such that “All you see is all there is”, causing people to make erroneous judgments. Propaganda posts that contain solely factual information cannot be determined as “fake” by algorithms. The idea that algorithms will solve this problem is silly and foolish.
Social media is in quite a mess at this point and there is at this time,  no clear path forward to fix this. It may not be fixable.