Shock for Movie Studios: Authors Face Setback in Crucial AI Lawsuit

A federal court has made a notable ruling on the use of copyrighted materials by AI companies, particularly Anthropic, which is backed by Amazon. The court described their practices as “transformative” and set a precedent that could influence future cases involving AI.

The main question is whether AI firms can claim fair use, which allows creators to build upon existing works without permission. In this case, the court sided with Anthropic, at least for training its AI. U.S. District Judge William Alsup stated that the technology could be among the most significant advancements we’ll witness in our lifetimes.

However, Anthropic isn’t off the hook. They face a trial for allegedly downloading seven million books to train their AI system, Claude. Just buying copies later doesn’t erase that initial illegal download. They may have to pay hundreds of millions, especially if Disney and Universal uncover similar issues with their AI, Midjourney.

In 2022, authors accused Anthropic of illegally copying their books for its chatbot. Instead of fighting the lawsuit, Anthropic aimed for a decision on fair use. The court found that authors can’t stop Anthropic from using their works to train its AI, much like they can’t prevent readers from learning from their books.

The court emphasized that reading and learning from texts is a normal part of writing. It argued that if a person reads classic literature and uses it as inspiration, that wouldn’t violate copyright. Similarly, Anthropic’s AI draws upon existing works to create something new rather than simply copying them.

This ruling may not be what Disney or Universal wanted. They are suing Midjourney, which creates images similar to existing movies. Alsup’s findings suggest that AI tools like Midjourney could be seen as producing entirely new works rather than infringing on copyrights.

Daniel Barsky, an intellectual property lawyer, argues that this decision will help AI developers argue for fair use when training their models. However, there are important distinctions: the authors didn’t claim that Anthropic violated their copyrights with the outputs of its AI. If they had, they would likely have lost that argument, as safeguards prevent infringing content from reaching users.

In contrast, Midjourney has faced criticism for sometimes creating near-identical images from popular films. For instance, when asked for an image of Thanos from “Infinity War,” the results often look strikingly similar to the official promotional material. Users have noted that the tool can mimic styles from various animated films, raising concerns about copyright infringement.

The core issues with Midjourney include potential unauthorized use of copyrighted films. In Anthropic’s case, the court found that illegally downloading books for training purposes is not acceptable under fair use. The penalty for such piracy can be significant, with statutory damages reaching up to $150,000 for each infringed work.

As AI technology continues to evolve, these legal battles will shape how creators and companies navigate copyright laws. The discussion is ongoing, and both sides will likely keep pushing for their rights in this rapidly changing digital landscape.

Source link

anthropic