The AIs can reference and resource check, essentially creating a network of cross-references that can categorize based on likelihoods. They can look at millions of data points and apply a confidence score, based on all kinds of bits of information as to whether something is likely to be true or not. There are lots of ways for AI to categorize and evaluate content, as well as identify sources of Gen-AI. I also suspect that more and more people will opt-in to human-built content, and people will want "proof of human" checks, providing more data points.
Journalism has been attacked on multiple fronts and that truly only benefits billionaires and corrupt politicians.
Perhaps it always was, but "back in the day" the consumer was more interested in getting useful information, but now we are just looking to be entertained.
Oh. I didn't know AI could do that. I'd be super curious to see how accurate those confidence scores get.