"With regard to the use of copyrighted materials to train LLMs, Judge Alsup described the use as "transformative-spectacularly so," which bears on the first fair use factor. The opinion analogizes model training to human reading and learning. That purpose, the court held, is fundamentally different, i.e., transformative, from the authors' purpose in writing books to be read for entertainment or education purposes."
"The same order blessed Anthropic's "print-to-digital" process as fair use. Anthropic bought physical copies of books, scanned them, destroyed the paper copies, and stored searchable digital versions for internal use. The opinion drew a sharp line, however, at Anthropic's acquisition of millions of pirated books downloaded from shadow libraries. He viewed the acquisition of these books and the training of the LLM based on these books to be distinct steps, each of which would be evaluated with regard to fair use."
Three federal district court decisions in 2025 began to outline fair-use boundaries for AI training on copyrighted works. Bartz v. Anthropic addressed authors' claims about use of lawfully acquired and pirated books to train a large language model. Judge Alsup characterized model training as "transformative—spectacularly so" and compared it to human reading and learning, finding a fundamentally different purpose from authors' aims. The court approved a print-to-digital scanning and internal storage workflow as fair use. The court nevertheless treated acquisition of pirated books as a separate step and found that procurement was not fair use on the record. The decisions show divergent fair-use outcomes and signal a need for appellate clarification.
Read at IPWatchdog.com | Patents & Intellectual Property Law
Unable to calculate read time
Collection
[
|
...
]