A display-first deal that rewrites the AI content playbook
The Getty Images OpenAI deal is a multi-year display-focused licensing agreement that lets ChatGPT Search show Getty’s licensed photos in its answers while keeping model-training rights, financial terms, and creator compensation undisclosed, signalling a move from unlicensed scraping to structured AI image licensing that investors love but creators cannot yet fully judge. This is not a neutral experiment; it is a statement about who gets paid, who stays in the dark, and how AI search will be built. Getty’s stock price briefly surged more than 200% in premarket trading after the announcement, before settling at gains of around 108% during regular trading, as investors rushed to price in a new AI-driven licensing model. For users, the promise is richer ChatGPT Search images; for the industry, the message is that content owners will no longer accept a “free-for-all” data grab.
What users gain: richer ChatGPT Search images with strings attached
For everyday users, the upside is obvious: ChatGPT Search images can now include professional, licensed photographs from Getty’s vast archive, adding detail and credibility to visual answers. Getty says its content will appear directly in ChatGPT, “significantly improving the richness of visual responses” by folding high-quality imagery into the search and discovery experience. Agreement terms allow ChatGPT to show Getty imagery in-line with answers, with visible credits and source links expected to explain where pictures come from and how they are licensed. That matters because AI search is becoming a primary interface for information; if images are anonymous or unlicensed, trust erodes fast. Yet users should not confuse display rights with training rights. What they see on screen is tightly controlled, but what happens behind the scenes—whether these images ever help train future models—remains unspoken and therefore contested.
Why Getty pivoted from lawsuits to AI image licensing
This deal is less a sudden embrace of AI and more a forced evolution. Getty previously sued an image-generation company over alleged scraping of more than 12 million photos for model training, a case that exposed the limits of current copyright law and showed how hard it is to stop unlicensed AI use through courts alone. At the same time, Getty’s share price had dropped more than 50% this year amid fears that AI tools could undercut demand for traditional stock photography. In that context, the OpenAI agreement is rational: having tested the courtroom route and found it wanting, Getty appears to have concluded that licensing is the surer path to getting paid. The company has already opened its archive to another AI search service via content licensing agreements, reinforcing that this is now a deliberate strategy, not a one-off experiment.
Creator rights in AI: a precedent with missing pieces
The most uncomfortable part of the Getty Images OpenAI deal is what it does not say. Getty works with nearly 600,000 content creators and hundreds of content partners, covering imagery from more than 160,000 events each year, yet neither Getty nor OpenAI has disclosed contributor economics or any separate payment for AI training rights. As a result, creators are asked to celebrate a stock-market rally they do not yet know how they share in. One quotable fact from the announcement sums up the tension: “commercial terms remain private because neither company has disclosed pricing, contributor economics, or any separate training-rights payment.” If AI search becomes a major revenue stream, photographers will demand clear answers on attribution, opt-outs, modification limits, and whether their work feeds future models without specific consent. Until those answers exist, this deal is a partial win for creator rights in AI, not a complete one.
Stock markets cheer, but the AI content licensing fight is far from over
Investors clearly see this content licensing agreement as a template for how legacy archives can earn money from AI search. Getty shares closed at $1.35, up 123% from $0.61 after the deal became public, in what investors treated as a test of whether Getty can turn its library into AI-search revenue. Across publishing, music and now stock photography, licensing deals are emerging as the preferred answer to the AI question: content owners rent access instead of watching their material scraped for free. At the same time, the OpenAI agreement explicitly focuses on display and stops short of allowing the images to train OpenAI’s own generator, DALL·E, preserving Getty’s stance against broad training use. This balance—licensed display, ambiguous training—will shape the next wave of AI content deals. The real test is whether future agreements can align user value, investor enthusiasm, and creator rights instead of forcing a trade-off between them.
