This is so to the point: OpenAI Furious DeepSeek Might Have Stolen All the Data OpenAI Stole From Us (404media.co), thanks to Y. C. for sharing.
The question: if Meta and OpenAI are allowed to steal copyrighted content and make money out of it and out of any legal consent, and if all the LibGen/ZLib/Anna's Archive content was sniffed by AI models it means each of those big tech companies consider it is legal to steal content and grow their business without respect for the law. I would assume so that if this kind of act of piracy is treated as necessary for businesses, piracy is de facto legalized and commoditized.
They are all pirating content directly or indirectly by using AI models or libraries that make this content accessible to any mere mortal and for every business.
I do agree with Anna & their team, this is a clear evolution of the jurisprudence and a copyright reform is therefore needed. Until then, there is no ethical distinction between big tech and "pirates". See also Copyright reform is necessary for national security (annas-archive.li)