This: it won't happen immediately and I'd go even further to say that it even if trust in images drops to zero, it's still going to generate a lot of hell.
I've always been able to say all sorts of lies. People have known for millennia that lies exist. Yet lies still hurt people a ton. If I say something like, "idle_zealot embezzled from his last company," people know that could be a lie (and I'm not saying you did, I have no idea who you are). But that kind of stuff can certainly hurt people. We all know that text can be lies and therefore we should have zero trust in any text that we read - yet that isn't how things play out in the real world.
Images are compelling even if we don't trust that they're authentic. Hell, paintings were used for thousands of years to convey "truth", but a painting can be a lie just as much as text or speech.
We created tons of religious art in part because it makes the stories people want others to believe more concrete for them. Everyone knows that "Christ in the Storm on the Sea of Galilee" isn't an authentic representation of anything. It was painted in 1633, more than a century and a half after the event was purported to have happened. But it's still the kind of thing that's powerful.
An AI generated image of you writing racist graffiti is way more believable to be authentic. I have no reason to think you'd do such a thing, but it's within the realm of possibility. There's zero possibility (disregarding supernatural possibilities) that Rembrandt could accurately represent his scene in "Christ in the Storm on the Sea of Galilee". What happens when all the search engine results for your name start calling you a racist - even when you aren't?
The fact is that even when we know things can be faked, we still put a decent amount of trust in them. People spread rumors all the time. Did your high school not have a rumor mill that just kinda destroyed some kids?
Heck, we have right-wing talking heads making up outlandish nonsense that's easily verifiable as false that a third of the country believes without questioning. I'm not talking about stuff like taxes or gun control or whatever - they're claiming things like schools having to have litter boxes for students that identify as cats (https://en.wikipedia.org/wiki/Litter_boxes_in_schools_hoax). We know that people lie. There should be zero trust in a statement like "schools are installing litter boxes for students that identify as cats." Yet it spread like crazy, many people still believe it despite it being proven false, and it has been used to harm a lot of LGBT students. That's a way less believable story than an AI image of you with a racist tattoo.
Finally, no one likes their name and image appropriated for things that aren't them. We don't like lies being spread about us even if 99% of people won't believe the lies. Heck, we see Donald Trump go on rants about truthful images of him that portray his body in ways he doesn't like (and they're just things like him golfing, but an unflattering pose). I don't want fake naked images of me even if they're literally labeled as fake. It still feels like an invasion of privacy and in a lot of ways it would end up that way - people would debate things like "nah, her breasts probably aren't that big." Words can hurt. Images can hurt even more - even if it's all lies. There's a reason why we created paintings even when we knew that paintings weren't authentic: images have power and that power is going to hurt people even more than the words we've always been able to use for lies.
tl;dr: 1) It will take a long time before people's trust in images "drops to zero"; 2) Even when people know an image isn't real, it's still compelling - it's why paintings have existed and were important politically for millennia; 3) We've always known speech and text can be lies, but we regularly see lies believed and hugely damage people's lives - and images will always be more compelling than speech/text; 4) Even if no one believes something is true, there's something psychologically damaging about someone spreading lies about you - and it's a lot worse when they can do it with imagery.
I've always been able to say all sorts of lies. People have known for millennia that lies exist. Yet lies still hurt people a ton. If I say something like, "idle_zealot embezzled from his last company," people know that could be a lie (and I'm not saying you did, I have no idea who you are). But that kind of stuff can certainly hurt people. We all know that text can be lies and therefore we should have zero trust in any text that we read - yet that isn't how things play out in the real world.
Images are compelling even if we don't trust that they're authentic. Hell, paintings were used for thousands of years to convey "truth", but a painting can be a lie just as much as text or speech.
We created tons of religious art in part because it makes the stories people want others to believe more concrete for them. Everyone knows that "Christ in the Storm on the Sea of Galilee" isn't an authentic representation of anything. It was painted in 1633, more than a century and a half after the event was purported to have happened. But it's still the kind of thing that's powerful.
An AI generated image of you writing racist graffiti is way more believable to be authentic. I have no reason to think you'd do such a thing, but it's within the realm of possibility. There's zero possibility (disregarding supernatural possibilities) that Rembrandt could accurately represent his scene in "Christ in the Storm on the Sea of Galilee". What happens when all the search engine results for your name start calling you a racist - even when you aren't?
The fact is that even when we know things can be faked, we still put a decent amount of trust in them. People spread rumors all the time. Did your high school not have a rumor mill that just kinda destroyed some kids?
Heck, we have right-wing talking heads making up outlandish nonsense that's easily verifiable as false that a third of the country believes without questioning. I'm not talking about stuff like taxes or gun control or whatever - they're claiming things like schools having to have litter boxes for students that identify as cats (https://en.wikipedia.org/wiki/Litter_boxes_in_schools_hoax). We know that people lie. There should be zero trust in a statement like "schools are installing litter boxes for students that identify as cats." Yet it spread like crazy, many people still believe it despite it being proven false, and it has been used to harm a lot of LGBT students. That's a way less believable story than an AI image of you with a racist tattoo.
Finally, no one likes their name and image appropriated for things that aren't them. We don't like lies being spread about us even if 99% of people won't believe the lies. Heck, we see Donald Trump go on rants about truthful images of him that portray his body in ways he doesn't like (and they're just things like him golfing, but an unflattering pose). I don't want fake naked images of me even if they're literally labeled as fake. It still feels like an invasion of privacy and in a lot of ways it would end up that way - people would debate things like "nah, her breasts probably aren't that big." Words can hurt. Images can hurt even more - even if it's all lies. There's a reason why we created paintings even when we knew that paintings weren't authentic: images have power and that power is going to hurt people even more than the words we've always been able to use for lies.
tl;dr: 1) It will take a long time before people's trust in images "drops to zero"; 2) Even when people know an image isn't real, it's still compelling - it's why paintings have existed and were important politically for millennia; 3) We've always known speech and text can be lies, but we regularly see lies believed and hugely damage people's lives - and images will always be more compelling than speech/text; 4) Even if no one believes something is true, there's something psychologically damaging about someone spreading lies about you - and it's a lot worse when they can do it with imagery.