Anthropic hathas received mixed results in a class-action lawsuit brought by a number of authors who claimed that the company used their work without permission. William Alsup, senior district court judge of the US District Court for the Northern District of California, determined that Anthropic’s training of its AI tool on copyrighted work was protected as fair usage.
The development of large language models for artificial intelligent has created a copyright boondogle, as creators try to protect their works while tech companies skirt the rules and look for loopholes in order to gather more training material. Alsup’s decision is likely to set legal precedents for what AI tools are allowed and not allowed to do.
Fair use can be determined if the output of copyright materials is “transformative,” or does not substitute the original work. “The technology at issue was among the most transformative many of us will see in our lifetimes,” Alsup wrote.
Despite the fair use designation, the ruling does still provide some recourse for the writers; they can choose to take Anthropic to court for piracy. “Anthropic downloaded over seven million pirated copies of books, paid nothing, and kept these pirated copies in its library even after deciding it would not use them to train its AI (at all or ever again),” Alsup wrote. “Authors argue Anthropic should have paid for these pirated library copies. This order agrees.”
