One legal way to share news stories which generate controversy is to have an AI make a summary of the article, rewording it to remove any copyright from the article (this legal trick only works when the news site in question is reporting on factual events, since facts can’t be copied as per Feist vs. Rural 1991, and I make sure that the summary has “a different sequence, pace, and setting than the original article” as per legal judgments on whether an AI summary of factual events as reported by a news source is subject to copyright).
Then it's no longer a record.
It might be useful for your own memory since you can decide that you trust what was written. IE you trust that when it was written, you agreed at the time that it was an accurate summary, and later you only have to trust your earlier self.
But "My ai said that the site said ~essentially~ this on this date." is useless for anything else.
But it suggests an interesting idea. Maybe if we had some ai equivalent of trusted vetted accountable court reporters who write down notes and draw sketches in rooms where cameras aren't allowed.
That is purely second-hand reporting that we more or less trust, partly because there is usually more than one of them from competing venues reporting on the same event, partly because they are people with something to lose if they are unfaithful or even merely inept, partly because the facts are ultimately vettable with some effort for those who really need to.
There's no such ai yet. Even if one exists that could do the job, there is no consensus or basis for it, for everyone to recognize it as a trustworthy source of second hand reporting. But maybe if there were more than one and operated by different camps, they could grow to become recognized as a valid source of info.