It seems increasingly clear that the companies building LLMs are engaged in massive copyrights and intellectual property rights violations, trade-mark infringements, and, perhaps, patents when they are training their LLMs on massive databases of texts, images, and videos (etc.).