It's possible that AI will make lies and propaganda easier, but we're already seeing plenty of people falling for obvious lies and clickbait.
As a simple example, just look at how in the USA, inventor of the internet, QAnon has gained zealous, violent supporters, over a bunch of intentionally cryptic messages and literal keyboard mashing to form a "code".
Look at the Reddit frontpage for an example of how easily people will believe the supposed context provided with a video. It usually takes hours for someone to actually check and comment "actually, this is a video from 2009 from another country" somewhere way down the bottom of the discussion thread. In private chat groups, this type of checking is almost entirely absent.
Why bother using AI when you can take an old video, add a title referring to a recent event, and spreading it around on social media? Good AI is expensive, lies are cheap.
As a simple example, just look at how in the USA, inventor of the internet, QAnon has gained zealous, violent supporters, over a bunch of intentionally cryptic messages and literal keyboard mashing to form a "code".
Look at the Reddit frontpage for an example of how easily people will believe the supposed context provided with a video. It usually takes hours for someone to actually check and comment "actually, this is a video from 2009 from another country" somewhere way down the bottom of the discussion thread. In private chat groups, this type of checking is almost entirely absent.
Why bother using AI when you can take an old video, add a title referring to a recent event, and spreading it around on social media? Good AI is expensive, lies are cheap.