BREAKING: Judge rules Anthropic did not violate copyright law

Issue 63

This seemed important enough to send a “breaking news” message:

Anthropic’s use of copyrighted works is deemed fair use

A federal judge has ruled that Anthropic’s use of copyrighted works to train AI is “exceedingly transformative” and fair use.

An important caveat is that in this particular case, the authors conceded that "training LLMs did not result in any exact copies nor even infringing knockoffs of their works being provided to the public."

The ruling says authors are still free to sue in the future if those things happen.

In a separate part of the case, Anthropic is still being sued over pirating millions of books to build a digital library and faces billions of dollars in damages.

Still, the ruling is being called the first of its kind and a major victory. The judge wrote that the activity of training was “like any reader aspiring to be a writer,” which is a framing that has been put forth by AI companies.

— CNBC, Wired

Quick hits

Libraries are supplying millions of public domain materials for AI training — Associated Press

Bloodhound Books partners with AI tool ProWritingAid to ‘enhance editorial and submissions process’ — The Bookseller

Chinese live streaming ‘AI clone’ outsells the human live streamer (but the audience was told it was AI, and I wonder if the novelty led to increased sales) — ChoSun Biz

Axel Springer Aims to Boost Value With AI, Calls Time on Clicks-and-Ads Model — Wall Street Journal

^{Written by a human}