Comment by smcg

you massively collapsed what AI companies have been doing by comparing it to old internet-scraping. Facebook flat-out admitted that they scanned copyrighted books for their AI. The image generators most definitely trained on copyrighted images.