That would be in the far future if even. Right now there are many ways to identify an AI generated picture especially if it has people in it (the infamous hands problem). And there's software that can reliably tell, too.
It's, as always, not a question about the tech but how we deal with it. If we were all reasonable and questioned such pictures if they pop up somewhere, you'd be right. However, if just enough people believe them to be real, the damage is done, even though if it proven to be fake afterwards.
That is most likely the reason blackmail and believing fakes are real didn't die and never will. Too many people believe what they want to believe, no matter what the actual truth is.