Getty Images and OpenAI announced a multi-year display partnership on June 22, 2026 that will bring Getty’s licensed photo and editorial library directly into ChatGPT search results. The deal covers over 400 million assets, including premium editorial content from sport, entertainment, and news coverage.
The market reaction was immediate. Getty Images shares surged over 140% in trading following the announcement, one of the largest single-day moves in the stock’s history. Investors read the deal as confirmation that content licensing, not litigation, is the model that wins in the AI era.
What the Deal Actually Does
Under the terms of the partnership, Getty content will appear in OpenAI’s search and discovery experiences inside ChatGPT. When users search for visual information (a news event, a sporting moment, a business concept), ChatGPT will be able to surface licensed images alongside its text responses.
This is a display partnership, not a training deal. As of the announcement, Getty has not authorised its images to be used for AI model training. The distinction matters. Getty has maintained a more cautious position on training rights than on display rights, and this deal preserves that line.
Getty CEO Craig Peters framed it clearly: “High-quality, licensed visual content makes AI-powered search and discovery more useful and more trustworthy. This partnership reflects a shared recognition of that, and together we will deliver richer visual experiences to ChatGPT users.”
A Pattern Worth Watching
The Getty deal is not the first of its kind. It joins earlier display partnerships between Getty and Perplexity, and between other content owners and AI platforms seeking to offer richer, more reliable search results. But the scale here is different. Four hundred million assets, including editorial photographs that cover decades of world news, represent a significant upgrade to what AI search can return.
For OpenAI, the deal solves a real product problem. Text-only search results feel thin when users ask about visual topics. Licensed imagery, clearly attributed, is a better answer than either leaving images out or generating images from scratch for factual queries.
For Getty, it validates a licensing revenue model at a time when many content businesses are still figuring out how to exist alongside AI.
What This Means for Business
If you run a business that creates or distributes content, this deal is a signal about where the AI content ecosystem is heading, and it has practical implications.
Content attribution is becoming automated. As AI platforms integrate licensed content libraries, the systems tracking what content was used, where it came from, and what licensing applies are becoming more sophisticated. Businesses that already manage content provenance well will adapt easily. Those treating AI-generated content as a copyright-free zone are building risk.
Licensed content commands a premium again. For years, the economics of online publishing pushed toward cheap or free stock images. The AI era is reversing that. Quality, licensed, clearly-attributed imagery is worth something to platforms that need to be defensible on copyright grounds. That is good news for professional content creators.
AI-search visibility is a new distribution channel. When ChatGPT can surface licensed photos in search results, the images that are discoverable inside AI platforms become a distribution consideration alongside traditional SEO. If you create visual content — product photography, editorial images, infographics. How that content gets indexed and licensed for AI display will matter for reach.
For agencies and marketers specifically: Understand what visual content your AI tools are actually pulling from. If you’re using ChatGPT search for research that then informs visual creative decisions, knowing those images come from licensed Getty stock rather than scraped-and-unattributed sources is a meaningful risk reduction.
The Bigger Picture
This deal lands at a moment when AI platforms are moving from training-focused content acquisition to building sustainable content supply chains. Training models requires access to data. Deploying those models in products that people trust every day requires something different: licensed, attributed, defensible content that content owners are actually happy to see used.
Getty’s willingness to partner rather than litigate (after years of taking a harder line) reflects that the industry has found a workable commercial framework. More deals like this will follow.
For businesses building with or alongside AI, the lesson is straightforward: the IP environment around AI-generated and AI-surfaced content is stabilising, not disappearing. Getting your content licensing practices in order now is less expensive than cleaning them up later.