Getty Sued AI. Now It’s Selling To AI, While Dodging The One Question That Counts.

Getty Images, the stock photography giant that sued Stability AI for copyright infringement in 2023, has now struck a licensing deal with OpenAI—turning ChatGPT into a storefront for its 600,000 contributors' photographs. The company remains conspicuously silent on the one question that matters most: whether those same images will also be used to train OpenAI's models.

Getty's about-face—from suing an AI company to partnering with one—signals a pragmatic shift in the stock photo industry's relationship with generative AI. The deal allows ChatGPT users to purchase licensed Getty images directly through the chatbot, effectively making OpenAI a distribution partner. But critics worry this opens the door for training data extraction without contributor consent.

Getty first sued Stability AI in February 2023, alleging that Stability copied over 12 million images from Getty's database to train its Stable Diffusion model. That lawsuit, along with similar actions by other artists and stock agencies, aimed to establish a legal precedent that AI companies must license training data. Now, the company is negotiating a license for its copyrighted content—but only for commercial output, not for training.

Key details remain sparse. Neither Getty nor OpenAI has disclosed the financial terms of the partnership, nor whether Getty contributors will receive additional compensation if their works are used in model training outside the licensing feed. Getty's refusal to address the training question has alarmed many of the 600,000 photographers and illustrators who supply its library. In a statement to Forbes, a Getty spokesperson said the company "remains committed to protecting creator rights" but declined to comment on training provisions.

Industry analysts view this deal as a strategic hedge. By licensing output, Getty secures a new revenue stream and positions itself as a safe, authorized source for AI-generated imagery. Meanwhile, it keeps its options open in ongoing litigation against other AI firms. But the silence on training data suggests a delicate balancing act: openly addressing it could undermine Getty's lawsuits or alienate contributors.

The broader implication is that copyright disputes around AI training data are far from resolved. Getty's deal sets a precedent—but only for the front end. The back end, where models learn from billions of images, remains a legal and ethical minefield. Regulators and legislators are watching closely, especially in the EU and US, where cases like Getty v. Stability could shape future AI copyright law.

What happens next? Contributors may push for transparency or collective action. Other stock agencies like Shutterstock, which already has a similar deal with OpenAI, may face renewed scrutiny. Getty's legal battle with Stability AI continues, and the outcome could determine whether training-data licensing becomes mandatory. For now, Getty is selling to AI—but it hasn't answered the question its contributors are asking.