Getty Images spent two years trying to take down AI image generators in court. This week, it joined one.
Getty announced a display partnership with OpenAI on Monday that puts its licensed photos directly inside ChatGPT search results, with attribution and links back to Getty's contributors. Wall Street loved it. Shares jumped more than 120% in early trading and at one point were up nearly 200%, according to Bloomberg. For a company that had become the public face of the "AI is stealing from creators" lawsuit, that's quite the reversal.
The pivot makes more sense when you look at how Getty's actual lawsuit went. In its case against Stability AI in the UK, Getty quietly dropped most of its biggest claims during the trial. It abandoned its training claim after acknowledging there was no evidence that Stable Diffusion was trained on its images in the UK. It abandoned its outputs claim. By the time the High Court ruled in November, the judgment noted that Getty itself was no longer even arguing Stable Diffusion stored copies of its work, with the court observing that "Getty Images do not say that Stable Diffusion is itself a copy of, or that it stores within it any copies of, the Copyright Works."
Translation: the company that became famous for fighting AI in court had quietly lost most of the fight before the judgment came down. Suing wasn't working, so licensing became the play.
Now compare Getty's path to Shutterstock, which figured this out years ago. While Getty was suing, Shutterstock signed a six-year licensing deal with OpenAI in 2023 to provide images, video, music and metadata for training. In March of this year, Shutterstock's SVP of Data Licensing Daniel Mandell explained the logic in a way most stock libraries are now coming around to: AI models aren't static, they have to be continuously retrained, which means high-quality licensed content is "the crown jewel."
That's the part most people miss about training data. It isn't a one-time sale. Models need to keep being refined, which means anyone who owns a large, well-labeled archive of human-made content has a recurring revenue stream sitting there if they want it. As The Next Web put it, the company that tried hardest to fight AI in court has decided the more durable position is to license to it.
Getty isn't alone in landing on that conclusion. Earlier this month, the National Music Publishers' Association announced industry-wide licensing deals with AI music startups Udio and Suno. Its CEO David Israelite framed the new posture pretty cleanly: "Litigating against bad AI actors and licensing good AI partners is not in conflict. NMPA will do both. And for companies that don't take this approach, you know it's coming."
For Getty specifically, the OpenAI deal gives it a foothold in what is starting to look like the next big distribution channel for images, with people increasingly asking a chatbot for a photo instead of running a Google search. CEO Craig Peters argued that licensed content makes AI search "more useful and more trustworthy," which is the kind of thing a CEO has to say when announcing a deal with the company your industry was supposed to be at war with.
The "AI is theft" narrative was always going to soften the moment one of its loudest critics found a way to get paid, and Getty just made the math obvious for everyone else holding a content library. Suing AI labs is a strategy with a ceiling. Licensing to them is a strategy with a recurring revenue line. Expect more publishers, labels, and stock archives still publicly bashing AI to quietly cut similar deals by the end of the year. The lawsuits won't stop, but they're going to start looking a lot more like opening bids than principled stands.
