In one of the first rulings of its kind, a federal judge ordered AI company Anthropic to pay $1.5 billion to a group of writers whose works were used to train its models. The catch: the judge ruled the training itself was lawful - the penalty was for pirating books from illegal shadow libraries, not for using them to build AI.
The distinction matters for legal professionals watching copyright law collide with generative AI. Judge William Alsup compared an LLM's ingestion of trillions of words to a writer's study of literature.
"Like any reader aspiring to be a writer, Anthropic's LLMs trained upon works not to race ahead and replicate or supplant them - but to turn a hard corner and create something different," the judge wrote.
Attorney Cathy Gellis, who specializes in intellectual property and technology law, said the ruling actually favors AI companies. A $1.5 billion fine is modest for a company projecting about $200 billion in annual revenue by 2028.
"I think it is generally good news for AI training that he looked at what was going on and really sort of thought it analogous to reading a copyrighted work as opposed to copying a copyrighted work," Gellis said. "Copyright law hinges on copying, but it doesn't hinge on using the work or experiencing the work, consuming the work, reading the work."
The fair use question
Copyright law hasn't been updated since 1976. Judges are interpreting 50-year-old guidelines to answer questions that could shape the AI industry for decades.
"Everybody is very worried right now because the law is all over the place, and it's because of this question," said Jason Henderson, senior attorney and founder of the IP & Media Practice at JWL International. "They know that the AI model has been trained on so much stuff, and the law has not really caught up to that question."
These cases hinge on fair use - whether using a copyrighted work is "transformative" enough to be legally permissible. Fair use protects commentary, criticism, parody, and education. Judges weigh the purpose of the work, the amount used, and its impact on the market.
"Copyright is always about protecting and growing the market," Henderson said. "The courts are kind of all over the place in their reasoning [in AI cases]. What's tending to win is if what you're doing is you're training on somebody's property because your purpose is to directly compete, then the courts will frown on it… If what you're doing is not going to compete, then the courts are tending to find ways that it will be okay."
Henderson points to Thomson Reuters v. Ross Intelligence, where a court ruled that Ross's use of Reuters' content to build a competing AI legal platform was not transformative. "Ross's use is not transformative because it does not have a 'further purpose or different character' than Thomson Reuters's," Judge Stephanos Bibas wrote. Authors have argued chatbots compete with them by generating synthetic books, but that argument has not yet prevailed in court.
AI-generated content and ownership
Gellis distinguishes between copyright in AI training and copyright in AI-generated output. In Thaler v. Perlmutter, a court ruled that a 100% AI-generated work is not copyrightable. That raises practical questions: how do you prove whether a work was AI-generated, and what percentage of human involvement is enough?
"If you write your novel in [Microsoft] Word and run spell check, we kind of feel comfortable with the idea of saying that Word does not own your novel," Gellis said. "[AI] is forcing us to look at a whole bunch of decisions that we kind of ignored for a while."
Most AI companies remain in pending litigation, so a definitive resolution is unlikely soon. "What you are seeing is that the initial opening volleys are being influential, and that influence itself could be undone if other courts decide different things, and it'll take later states of litigation to figure out which one will prevail," Gellis said. "But in the meantime, all these decisions are shaping everything that's happening. It would be kind of foolish for the AI companies to ignore them."
For legal professionals, the takeaway is practical: the doctrine of fair use is being tested in real time, and the outcomes depend heavily on whether an AI tool directly competes with the original work. Understanding these rulings matters for advising clients on AI for Legal risks, and for AI for Paralegals who may need to flag potential infringement issues in document review and discovery workflows.
Why this matters for legal professionals
Every copyright case involving AI training sets a precedent that will shape how courts treat future disputes. The distinction between reading and copying - between training and infringement - is now an active area of litigation with billions of dollars at stake. Legal teams advising AI companies, publishers, or individual creators need to track these rulings closely, because the law is being written case by case, and the early decisions are already influencing how AI companies approach data sourcing and licensing.
Your membership also unlocks:
