I don’t think there’s anything inherently wrong with using artificial intelligence to quickly whip up images of Donald Trump getting arrested in New York City, as long as those images are clearly labeled as being AI-generated. Such images certainly constitute protected speech under the First Amendment. But they’re realistic-looking enough to fool the casual observer, and you can see how someone who either wanted to make mischief or sow confusion could use images like this — Hillary Clinton getting arrested, Joe Biden wandering the White House rose garden in his bathrobe, Ron DeSantis waving a gun at Disney World — and attempting to create widespread erroneous perceptions about events that never happened.
A particularly nefarious person could create images of some sort of terrorist attack or disaster and then attempt to spread the rumor that the government and media are covering the event up. We’ve been able to edit photographs for a long time now, but AI is going to make faking photos easier than ever. Technology is rapidly outpacing humanity’s maturity and responsibility.