So when the various AI start publishing questionable stuff and add the published stuff into each others corpus to draw on, they all learn from each other to gen

So when the various AI start publishing questionable stuff and add the published stuff into each others corpus to draw on, they all learn from each other to generate even more nonsense? Eventually you wont be able to trust information from any unverifiable/non-human source and information-technology will die due to mistrust. When we were meant to automate the boring stuff, this isn't the end-point we were hoping for.