Ignore the Copyright Suit—Watch the Precedent Instead

The Seattle Times and Newsday are suing OpenAI for using their reporting to train an AI model. On the surface this looks like straightforward copyright infringement — news organizations didn't consent, didn't get paid, and now their work powers a commercial product. But the real story isn't what happened here. It's what happened eleven years ago in a courtroom that should have ended this argument before it started. In 2015, the Authors Guild sued Google Books with nearly identical claims: Google had scanned millions of copyrighted books without permission and used them to build a searchable index. The court found this transformative fair use. Not because the books were analyzed differently. Because computational analysis itself — running copyrighted text through an algorithm to extract meaning, qualified as transformation. The Authors Guild lost. Nothing has legally changed since then.

OpenAI's lawyers will cite that precedent in their opening motion. The newspapers will need something more than an indignant press release to get around it. Fair use doctrine doesn't care about the plaintiff's feelings or even the defendant's scale. It cares about whether the secondary use is transformative. Whether it does something new with the source material rather than simply copying it for the same purpose. Google didn't republish the books. OpenAI didn't republish the articles. Both extracted meaning computationally from copyrighted text to build something that didn't exist before.

The only viable legal theory these newsrooms have is arguing that training an AI chatbot is categorically different from building a search index. That one transformation counts and the other doesn't. That's not a clean argument. It's asking courts to draw a line through settled precedent and say that algorithmic analysis was fair use when Google did it but isn't when OpenAI does it, with no principle to explain the difference except that one harms newspapers and one didn't. The doctrine makes no distinction between kinds of harm. When you're building a defense around overturning precedent, you've already lost the easier fight.

What these suits really reveal is that the Authors Guild negotiated away everyone else's negotiating power. Once a court ruled that computational training on copyrighted text was transformative fair use, every AI company got a template. The newspapers coming forward now aren't fighting OpenAI. They're fighting a judicial decision from an era when the stakes didn't feel immediate. The precedent is the problem. The lawsuit is just the thing that makes us finally notice it.