Even though I haven't read the actual ruling, speculatively, it seems like a dangerous precedent. For instance, search engines and LLM-powered search might now be liable too, including with respect to training data. Science too, maybe.