According to a report, artificial intelligence can produce political propaganda almost as persuasive as Russian or Iranian fake news (though this may be a low bar):

In mid-April, Stamos and Stanford’s Shelby Grossman led a webinar where they previewed a forthcoming paper, “Can AI Write Persuasive Propaganda?” Their answer was a firm “yes.” AI was almost as persuasive as real Russian and Iranian propaganda used in recent years to distract, divide, and deceive Americans. . . .

Dishonest humans are already able to script claims that get bad scores from fact-checkers and create social media personas and accounts to spread misinformation, via ads, texts, tweets, hashtags, comments, and e-mail. But the Stanford study previews a far more disturbing future: authoritarians using AI to fabricate massive amounts of believable and manipulative longer-form content that looks no different than the news on a smartphone. We could be facing a world where it is very hard to figure out what is or is not true.