So the solution is to allow the AI scraping and hide the content, with significantly reduced fidelity and accuracy and not in the original representation, in some language model?

Don't forget the onslaught of ads that will distort the actual publications even more going forward.