That was to be expected. Chatbot-spam gets harder and harder to detect (and prove), and if Mods don't ban them on the spot for plagiarism, there is no limit to the amount of verbal diarrhea they can create.
OK well I'm in favor of banning on the spot, if they already have at least 3 posts deleted as AI spam yet continue to engage in AI spamming.
And yes, there are even more accounts which tweak the text a bit in order to make it appear they aren't using ChatGPT. They accomplish this by doing stuff like adding, "I believe" at the start, unnecessary punctuation, or bits of text here and there that are out of character for ChatGPT.
Case in point. By itself, this post comes back as
0.3% Fake according to Sapling.ai (though 100% AI according to Hivemoderation):
It's clear that saving money has become harder for many especially with rising living costs and stagnant wages. People are working more but struggling to set aside any savings. There's a need for changes whether it's better job opportunities, affordable housing or other policies that can make things easier for everyone. It's a complex issue but acknowledging the problem is a step toward finding solutions right?
When a comma is added in the first sentence, and "right?" is removed in the last sentence, the result changes dramatically, to
100% Fake:
It's clear that saving money has become harder for many, especially with rising living costs and stagnant wages. People are working more but struggling to set aside any savings. There's a need for changes whether it's better job opportunities, affordable housing or other policies that can make things easier for everyone. It's a complex issue but acknowledging the problem is a step toward finding solutions. right?
So when does a shitpost become spam territory, and when does spam become plagiarism? This is what we (well, more the mods) will need to establish going forward, and not even ChatGPT can help us answer this question.