ChatGPT’s search experience is about to get a major visual upgrade. On June 21, 2026, Getty Images announced in New York that it has signed a multi-year display partnership with OpenAI, granting the AI company rights to surface Getty’s vast library of licensed photographs, illustrations, and editorial imagery directly within ChatGPT. The deal, described as a “multi-year display partnership,” marks one of the most significant commercial agreements between a traditional stock media giant and a leading generative AI platform. It aims to weave high-quality, rights-cleared visuals into the fabric of conversational search, rewriting how users discover and interact with images.

For the first time, ChatGPT users will see Getty-owned or Getty-represented images in response to queries that call for visual context—think historical events, celebrity portraits, or travel destinations. These images will appear with proper attribution and, crucially, under a licensing framework that addresses the legal and ethical quagmire that has long surrounded AI’s use of copyrighted material. OpenAI is not acquiring the images to train models; the deal is specifically for display, meaning the images are shown as search results or illustrative content, not ingested for generative purposes. This distinction is critical in an era where copyright lawsuits against AI developers have become routine.

The Anatomy of the Deal

While the full financial terms remain undisclosed, the agreement is structured as a revenue-sharing model. Getty will receive compensation based on usage metrics—likely a combination of flat licensing fees and per-impression payments. OpenAI, in turn, gains access to over 400 million assets, including premium editorial content from Getty’s collection of sport, entertainment, and news imagery. The deal also covers content from iStock, Getty’s budget-friendly subsidiary, broadening the range of visuals available to ChatGPT users.

The partnership is exclusive in the sense that it gives OpenAI a first-mover advantage among large language models. Other AI companies, such as Google and Meta, have struck their own image-licensing deals—Google with Shutterstock and Meta with Shutterstock and others—but the Getty-OpenAI pact is the first to bring a comprehensive editorial library directly into a conversational AI interface. This is not just about pretty pictures; it’s about authoritative visual information. When ChatGPT answers “Who won the 2024 Oscar for Best Actor?” it can now show a rights-cleared photo of the winner on stage, alongside the answer.

How It Works for Users

From a user perspective, the integration will be seamless. When a query triggers an image result, ChatGPT’s interface—whether on the web, mobile app, or desktop—will display thumbnails or full-bleed images, each with a Getty Images watermark and a clickable attribution link. Clicking the image will take users to the Getty website, where they can license it for commercial use. This cuts a path from discovery to transaction, potentially turning ChatGPT into a top-of-funnel tool for Getty’s licensing business.

For free-tier users, the images will likely be served with a Getty watermark, while paying subscribers to ChatGPT Plus or Pro might see higher-resolution versions without watermarks, possibly under a different licensing tier. OpenAI has not confirmed these details, but industry precedent suggests a tiered approach to balance accessibility with monetization.

The feature will roll out gradually across all ChatGPT platforms, including the Windows desktop app. Microsoft, OpenAI’s largest investor, has been integrating ChatGPT technology into Bing and Windows features like Copilot. While this deal is between OpenAI and Getty, the ubiquity of ChatGPT on Windows means millions of Windows users will soon see licensed images in their AI interactions. This could pave the way for similar integrations in Microsoft’s own Copilot experiences, though Microsoft has its own image-licensing agreements through Bing Image Search (via Shutterstock and others).

The Getty-OpenAI partnership is being hailed as a template for how AI companies can ethically use copyrighted visual content. Getty has been one of the most vocal advocates for creator rights in the AI era, notably winning an injunction in the UK against Stability AI for training its model on Getty images without permission. By opting for a display-only deal, OpenAI sidesteps the training controversy entirely. It does not need to scrape the web or rely on fair-use defenses; instead, it pays for access to a curated, legally clean library.

For photographers, artists, and agencies represented by Getty, the deal offers a new revenue stream. When their images appear in ChatGPT search results and drive licensing traffic, they receive a royalty. This model could become the standard for AI-powered search engines, balancing innovation with compensation. “This partnership signals that AI companies are willing to pay for premium content when it’s used to enhance user experience,” said Getty CEO Craig Peters—a stance that has been echoed in similar deals between news publishers and AI firms.

However, the devil is in the details. The exact payout structure to contributors is unknown, and some photographers have expressed skepticism on forums about whether the per-impression rates will be meaningful. There’s also the question of how often an image needs to be displayed before it generates significant income for its creator. Still, the alternative—having images used without any compensation—makes this a clear improvement.

The Competitive Landscape

OpenAI is not alone in seeking to legitimize visual search. Google’s Bard and Search Generative Experience (SGE) already pull images from licensed sources like Getty and Shutterstock, thanks to pre-existing agreements. Adobe Firefly, meanwhile, is trained entirely on Adobe Stock images, ensuring commercial safety. The difference with ChatGPT is the conversational context: images are not just search results but integral to a dialogue, making them more likely to be seen and clicked.

Shutterstock reaffirmed its own partnership with OpenAI in early 2026, but that deal is primarily for training data, not display. Shutterstock images power the DALL·E model that generates images within ChatGPT. The Getty deal, by contrast, is purely for surfacing real-world, non-generated visuals. This dual approach gives OpenAI a comprehensive visual toolkit: generative imagery for creative tasks, and licensed editorial imagery for factual queries.

Smaller stock platforms like Unsplash and Pexels, which offer free images under permissive licenses, are already integrated into some AI search tools. But these lack the editorial depth and brand recognition of Getty. For breaking news, celebrity photos, or archival footage, Getty remains the gold standard, and ChatGPT will now have a licensed conduit to that treasure trove.

Potential Pitfalls and Privacy Concerns

No deal of this magnitude comes without unease. Privacy advocates immediately raised flags about the data exchange between OpenAI and Getty. When a user clicks on an image, their referral data—potentially including the query that led to the image—could be shared with Getty. Both companies have stated that they will comply with global privacy regulations, but exactly what user information is passed remains unclear. Given OpenAI’s track record of iterating on data practices, this will be an area to watch.

Another concern is bias and editorial control. Getty’s editorial images are curated and captioned by human editors, which injects a layer of human judgment. When ChatGPT surfaces these images, it might inadvertently amplify certain news narratives or cultural perspectives over others. OpenAI has said it will use its safety systems to filter images that violate content policies, but those systems are imperfect. There’s also the risk that prominent placement of Getty images in ChatGPT results could crowd out independent photojournalists or local news sources that lack the resources to license their content to big platforms.

Furthermore, the deal does not cover the use of images for training future AI models. OpenAI has made it clear that the Getty content will not be used to improve its image-generation capabilities or to train any model. But as the line between search and generation blurs—consider ChatGPT’s ability to generate images via DALL·E and then refine them conversationally—the separation may become harder to maintain. If a user asks ChatGPT to “create an image like the one you just showed me of the Oscars,” what happens legally? OpenAI has not addressed such edge cases, and they could become flashpoints in future copyright discussions.

What It Means for Windows Users and the AI Ecosystem

For the millions of Windows users who have adopted ChatGPT as a daily tool, the Getty integration will be one of the most visible updates since the GPT-4o multi-modal launch. The Windows app, which already supports dark mode, file uploads, and voice conversations, will likely gain a dedicated image viewer pane or modal. This could make ChatGPT more competitive with traditional search engines, especially for visual research. Students, writers, and journalists using Windows to research articles will appreciate the context that a well-captioned Getty image provides.

The deal also underscores Microsoft’s broader strategy of embedding licensed content into AI experiences. While Microsoft has not explicitly partnered with Getty for Copilot, the two companies have a history of collaboration—Getty images have long been available in Microsoft Office and Teams via add-ins. It’s conceivable that future versions of Windows Copilot, which already pulls in web results, could mirror the ChatGPT integration under a shared licensing umbrella. Such a move would blur the lines between OpenAI and Microsoft’s AI offerings, but it would also give Windows an edge over macOS and ChromeOS in terms of native visual intelligence.

The Road Ahead

The multi-year timeframe suggests that both companies see this as a long-term play, not a trial balloon. As AI search evolves from text-based answers to multimedia-rich experiences, the value of a licensed image library will only grow. Competitors will be watching closely: Google, with its own image deals, may accelerate integration of similarly licensed content into its Gemini model. Apple, too, is reportedly exploring AI-powered visual search enhancements for Siri and Photos.

Getty has committed to expanding its AI partnerships, but it’s picky—it seeks deals that protect its contributors while extending its reach. An OpenAI exclusive could limit other platforms’ access, but Getty has historically played the field (it also supplies images to Amazon, Google, and Apple). The exclusivity here might apply only to direct conversational AI search, leaving room for other types of licensing.

One thing is certain: the days of AI scraping images indiscriminately are fading. As copyright norms are codified by courts and contracts, the AI industry is learning that sustainable growth requires paying for high-quality data. The Getty-OpenAI deal is not the first of its kind, but it is the most visible, and it sets a high bar for what responsible visual AI integration looks like.

In the end, users will be the ultimate beneficiaries. They’ll see richer answers, creators will get paid, and the legal cloud over AI images might finally lift—at least for those who choose to play by the rules. The partnership doesn’t solve every ethical dilemma of AI, but it proves that copyright and cutting-edge technology can coexist when both sides come to the negotiating table.