Anthropic gives court authority to intervene if chatbot spits out song lyrics

Anthropic did not immediately respond to Ars’ request for comment on how the guardrails currently prevent alleged jailbreaking, but the publisher appears satisfied with the current guardrails in accepting the deal.

Whether AI lyrics training infringes copyright is still unresolved

Now, the question of whether Anthropic has strong enough guardrails to prevent allegedly harmful output has been resolved, Lee wrote, allowing the court to focus on the issues surrounding “the publisher’s request in its motion for preliminary injunction that Anthropic not use the publisher’s unauthorized copy” controversy. Lyrics for training future artificial intelligence models. “

man expresses in his book Motion Against Preliminary Injunction This relief should be denied.

Anthropic’s court filing states: “Can generative AI companies use copyrighted content to train LL.M.s without a license? About two dozen copyright infringement cases are currently being litigated nationwide, but none are The case attempts to resolve this issue.

A spokesperson for Anthropic told Ars in a statement, “Claude was not designed for copyright infringement and we have a number of processes in place to prevent such infringement.”

“Our decision to include this provision is consistent with these priorities,” Anthropic said. “We continue to look forward to demonstrating that use of potentially copyrightable material in the training of generative artificial intelligence models is typically fair use under existing copyright law.”

The lawsuit could take months to fully resolve, as the question of whether artificial intelligence training is a fair use of a copyrighted work is complex and remains hotly contested in court. The stakes could be high for Anthropic, with losses that could trigger fines of more than $75 million and an order that could force Anthropic to disclose and destroy all copyrighted works in its training data.