One of the world’s largest photography agencies has struck a multi-year licensing agreement with OpenAI, a remarkable turn for a company that only months ago lost a high-profile copyright battle against another artificial intelligence developer.
Getty Images said the deal will see images from its library surface within ChatGPT’s search display, folding richer visual features into the chatbot. Crucially, the agreement stops well short of handing OpenAI the keys to Getty’s archive: it does not allow the images to be used to train OpenAI’s own image generator, Dall-E. Neither company disclosed the financial terms.
Investors liked what they saw. Shares in Getty Images jumped as much as 65 cents, or roughly 108 per cent, to $1.26 in early afternoon trading in New York, a rare burst of optimism for a stock that has been badly bruised this year.
Craig Peters, chief executive of Getty Images, framed the deal as a vote of confidence in licensed content over the free-for-all that has defined much of the AI land grab. “High-quality, licensed visual content makes AI-powered search and discovery more useful and more trustworthy,” he said. “This partnership with OpenAI reflects a shared recognition of that, and together we will deliver richer visual experiences to ChatGPT users.” The full terms were set out in Getty’s own announcement of the partnership.
Getty’s total catalogue stands at about 609 million images, placing it among the largest photography platforms on the planet. Yet scale has offered little protection from the market’s anxieties. The shares have shed more than 50 per cent of their value this year, dragged down by fears that AI-powered image generators could hollow out demand for traditional photo libraries altogether.
That existential worry helps explain Getty’s other big move: the company is finalising a $3.7 billion merger with its long-time rival Shutterstock. The two argue that combining forces will create a business with an unrivalled photo library and, just as importantly, the scale to invest in its own image-generation models rather than simply ceding the territory to Silicon Valley.
The OpenAI agreement marks a striking strategic pivot. In 2023, Getty sued Stability AI, the UK-based developer behind the Stable Diffusion model, alleging it had scraped more than 12 million images from Getty’s library without permission.
The case proved a cautionary tale about the limits of existing copyright law. Unable to prove that the training had taken place in the UK, Getty was forced back onto a secondary infringement claim, arguing that Stability had effectively imported an infringing product in breach of British copyright rules. A High Court judge found against the agency, ruling that Stability’s model did not actually store or copy images from Getty’s library. As Business Matters reported when the judgment landed, the decision was widely read as a setback for rights holders, even as parallel proceedings rumble on in the United States. Stability itself had earlier cast the lawsuit as an “existential threat” to the generative technology industry.
Having tested the courtroom route and found it wanting, Getty appears to have concluded that licensing is the surer path to getting paid.
It is not alone. Across publishing, music and now stock photography, the licensing deal has emerged as the industry’s preferred answer to the AI question. Getty signed a comparable agreement with Perplexity AI last November, while OpenAI has lined up deals with major publishers including News Corp, owner of The Times, as well as The Guardian and the Financial Times. The chatbot’s growing commercial ambitions are increasingly visible elsewhere too, from advertising trials within ChatGPT to its preparations for a stock market debut.
For Getty, the calculation is straightforward enough. If AI tools are going to reshape how people find and use images, the agency would rather be paid to be part of that future than litigate its way through it. As Bloomberg noted, the market reaction suggests investors, for now at least, agree.
