You might think Meta has scored a clean win in artificial intelligence and copyright law, but it’s not that simple. On June 26, Judge Vince Chhabria ruled in Meta’s favour in a lawsuit filed by 13 authors who claimed the company used their books to train its AI models without permission. According to the ruling, Meta is “entitled to summary judgment on its fair use defence” against the accusation of copyright infringement.
At first glance, this may look like a green light for companies like Meta to train AI using copyrighted materials. However, the judge clarified that this ruling shouldn’t be taken as proof that Meta’s actions were entirely legal. Instead, the win resulted from the authors’ legal team not making the proper case.
Why the case fell apart for the authors
In his decision, Judge Chhabria criticised the authors’ arguments. He said two of their key claims were “clear losers.” One was the idea that Meta’s Llama AI model could reproduce chunks of their books. The judge disagreed, stating that Llama isn’t capable of generating enough content from those books to be of any real concern.
The second weak point was the claim that Meta’s use of their work would harm their ability to license it for AI training. Judge Chhabria was blunt on this: the authors aren’t entitled to dominate the licensing market for training AI. He explained that while it may be a new and emerging area, the right to control that space doesn’t automatically belong to them.
The judge did admit that there was one potential argument—that AI models could flood the market with similar content, which might lower the value of the original works. However, even then, the authors failed to build a strong case. They didn’t provide enough evidence to show how Meta’s use of their work might lead to market harm.
What this means for AI training and fair use
This decision doesn’t mean the door is wide open for AI companies to train their models however they like. Judge Chhabria warned that the ruling should not be seen as a blanket approval of Meta’s practices. Instead, it highlights the need for well-argued, properly supported cases if copyright holders want to challenge AI companies effectively.
The ruling also followed a similar decision made just a day earlier. In that case, Judge William Alsup sided with Anthropic, another AI company, ruling that training AI on legally purchased books is considered fair use. Judge Chhabria pointed to this decision and noted that Judge Alsup had largely ignored the potential impact of generative AI on the market for the original works.
The legal world is clearly still figuring out where to draw the line regarding AI and copyright. While Meta came out on top this time, future challenges could have different outcomes, especially if plaintiffs come better prepared with stronger arguments.
