What’s on the Horizon for AI Copyright Litigation?

Here’s a revised version of the content with external promotional references removed or rephrased:

—

There are numerous copyright lawsuits currently working their way through the courts, with cases filed against all the major players—not only Anthropic and Meta but also Google, OpenAI, Microsoft, and others. On the opposing side, plaintiffs range from individual artists and authors to large organizations like Getty and the New York Times. The outcomes of these cases are poised to have a significant impact on the future of AI, determining whether model creators can continue accessing training data without charge. If not, they may need to enter into new licensing agreements or explore alternative methods for training their models, which could fundamentally alter the industry landscape.

That’s why last week’s victories for the technology companies hold significance. So, are all cases resolved? Not quite. A closer examination reveals that the rulings are not as straightforward as they may initially appear.

In both situations, a group of authors (with the Anthropic case being a class action involving 13 plaintiffs, including notable names like Sarah Silverman and Ta-Nehisi Coates) sought to demonstrate that a technology company infringed their copyright by utilizing their books to train large language models. In response, the companies contended that this training process qualified as fair use, a legal standard that allows the use of copyrighted works for specific purposes.

However, the similarities end there. Ruling in favor of Anthropic, senior district judge William Alsup expressed on June 23 that the firm’s use of the books was lawful because it was transformative, meaning it did not replace the original works but rather created something new. “The technology at issue was among the most transformative many of us will see in our lifetimes,” Alsup noted in his judgment.

In Meta’s case, district judge Vince Chhabria reached a different conclusion. He also sided with the technology company, but his focus was on whether Meta had negatively impacted the market for the authors’ work. Chhabria mentioned that he thought Alsup had underestimated the relevance of market harm. “The key question in virtually any case where a defendant has copied someone’s original work without permission is whether allowing people to engage in that sort of conduct would substantially diminish the market for the original,” he wrote on June 25.

Although the outcomes were the same, the reasoning behind them diverged, leaving ambiguity regarding their implications for other cases. On one hand, the rulings strengthen at least two iterations of the fair-use argument. On the other, there is some disagreement regarding the criteria that should be used to determine fair use.

Moreover, there are even more critical aspects to consider. Chhabria was explicit in his judgment that Meta’s victory did not indicate it was legally justified, but rather that the plaintiffs did not present a strong enough case. “In the grand scheme of things, the consequences of this ruling are limited,” he stated. “This is not a class action, so the ruling only affects the rights of these 13 authors—not the numerous others whose works Meta used to train its models. And, as should now be clear, this ruling does not establish that Meta’s use of copyrighted materials to train its language models is lawful.” This language suggests an open invitation for others with similar grievances to pursue their claims.

—

Is there anything more you’d like to adjust?