At this point I see no way for any social media company to talk about moderation (even legally mandated moderation) without getting dog piled. Note that even Musk is not advocating absolute free speech. In interviews about his attempt to purchase Twitter his frustration seems to be more about moderation being a black box not that moderation is not needed or wanted.
If you agree that AI driven moderation is "state of the art" and that moderation must be transparent and not a black box, how are you proposing that a company explain how transformers and hidden layers made a given moderation decision to the public when even the best PhDs can't really explain it to each other?