Meta's AI training under fire: Copyright Infringement allegations

In an increasingly tech-driven world, one of the hottest topics right now is the growing debate over the ethical and legal use of copyrighted material in the training of AI models. From public reports, it would appear that Meta has been caught with its hands in the copyrighted material cookie jar.

*Image credit: Wikimedia Commons

Background

Meta, the parent company of Facebook and Instagram, is facing a class action lawsuit for allegedly using pirated material to train its AI models, including LLaMA, without permission or proper licensing. The lawsuit claims that Meta used over 81TB of pirated eBooks from shadow libraries like Anna’s Archive, Z-Library, and LibGen to teach its AI how to respond to human prompts.

The situation escalated when court records revealed that despite internal concerns raised by Meta’s own employees about the use of pirated content, the company proceeded with torrenting (downloading) pirated books. This raises questions about company practices and the broader ethical considerations of using copyrighted material without a license.

What’s at stake?

This isn't the first time a major AI company has faced allegations of using unauthorized materials. OpenAI, Nvidia, and others have been similarly accused. For creators like authors, the unauthorized use of their works in AI training could be a serious violation of copyright law, and if the courts rule in favor of the plaintiffs, it could reshape how AI companies access and use data.

Why it matters

In addition to the ongoing legal battle, Meta has been criticized for scraping user data from Facebook and Instagram to train its AI models. The company admitted to using public posts and photos without users’ explicit consent, raising privacy and copyright concerns across the globe.

The bigger picture

This case highlights a significant issue in the intersection of AI, copyright, and data privacy. While companies argue that using public or pirated material falls under "fair use" or similar exceptions, the fact remains: creators' rights are often overlooked in the rush to advance technology.

As these cases progress, they could set important precedents for how the tech industry uses copyrighted material to fuel AI innovation without stepping on the toes of copyright holders.

This is just the beginning of a much larger conversation on intellectual property and AI.

Dr Bernard Dippenaar

Senior Associate | Intellectual Property Department

bdippenaar@ENSafrica.com