Will AI Settlements Spawn More?

A federal judge presiding over a class action earlier this summer said Anthropic made fair use of many authors' work to train its AI, but said the company violated copyright law by saving pirated books to a central library not necessarily used for AI training. That proved to be too many potential liabilities facing Anthropic, which just settled before an upcoming trial in December. Likewise, in the case against Eleven Labs, the defendant allegedly copied the voices of plaintiffs without consent and removed technical protections and copyright management notices from audiobooks in order to train their models. A stipulation staying the case was filed pending settlement. Legal scholars are wondering if these settlements facilitated by mediators signal how the industry will navigate the dozens of similar lawsuits pending nationwide. While settlement details remain confidential pending court approval, the timing reveals essential lessons for AI development and intellectual property law. With almost fifty other pending cases, Professor Edward Lee opines an increased probability of settlement (with which ChatGPT agrees), though they involve different companies, different types of works, various judges, and potentially enough variety in the training of the AI models to matter. Meta’s use of copyrighted works to train Llama was considered a fair use, even where Meta had obtained those works from piracy websites. However, the Meta decision was solely based on the record before the court and not a broad holding that all of Meta’s actions were fair use as to all possible plaintiffs, with other parts of the decision less favorable to the defendant. With those fair use precedents, some defendants in the other cases could feel they have decent defenses in their litigation. Anthropic was in "a unique situation," according to Professor James Grimmelmann, with as much as $1 trillion in piracy damages at stake in its worst-case scenario. "It's possible that this settlement could be a model for other cases, but it really depends on the details," he said to Reuters. Reportedly, Professor Chris Buccafusco was surprised Anthropic chose to settle, saying Anthropic was "in a position of decent strength" because of its fair-use determination, despite the piracy decision. The fate of the pending generative AI lawsuits could hinge on fair use, a still-evolving concept that no court had addressed in the cases until June. He also said Anthropic's settlement removes an early opportunity for a federal appeals court to consider fair use and issue a decision that would be binding on other cases and likely send the issue to the U.S. Supreme Court (where a bad precedent might ensue). See more here-- https://tinyurl.com/ydkynvmp and https://tinyurl.com/527phxkw and https://tinyurl.com/yn5rdruz and https://tinyurl.com/2z7nfwps and https://tinyurl.com/ayaswu8s