SAN FRANCISCO (AP) – A group of book authors have reached a settlement agreement with the artificial intelligence company Humanity after suing a chatbot maker for copyright infringement.
Both sides of the lawsuit “negotiated a proposed class settlement,” according to a filing by the federal court of appeals that said Tuesday the terms would be completed next week.
Humanity declined to comment Tuesday. Justin Nelson, the author's lawyer, said “historical settlements benefit all class members.”
In a major test case in the AI industry, a federal judge ruled in June that humanity did not break the law by training chatbot Claude on millions of copyrighted books.
However, the company was still on the hook and was brought to court for how they obtained those books by downloading them from the online “Shadow Library” of pirated copies.
San Francisco US District Judge William Alsup said in his June ruling that the AI system was distilled from thousands of written works. He said that because it is “typically transformative,” he could create his own texts of texts that are recognized as “fair use” under US copyright law.
“Like readers who are aiming to become writers, humanity (the AI-based language model) trained their works to not race, recreate, or replace them first.
The trio of Andrea Burtz, Charles Graeber and Kirk Wallace Johnson alleged in lawsuit last year that human practice amounted to “a massive theft,” and that the San Francisco-based company “seeks to benefit from the human representation and the human representation and talent behind each work.”
