AI search company Perplexity has signed a multi-year contract with Getty Images, providing Perplexity the right to display Getty’s library across its AI-driven search and discovery products. The contract symbolizes Perplexity’s shift to licensed media amidst ongoing concern about the source and attribution of AI responses. The alliance with Getty Images enables Perplexity software to present Getty Images photography and video in results, with credit and links to the original site.
This indicates that AI-generated summaries and answer cards can be shown alongside relevant editorial and inventory visuals, captions, and creator bylines, as opposed to unauthorized thumbnails and generic fillers. Perplexity notes that combining concise text-based responses with properly attributed visuals will contribute to result quality and transparency, while Getty executives see the deal as a crucial vote of confidence in AI products that utilize a conscience-based license process.
Why Getty Images matters for modern AI search
Getty’s extensive collection, which encompasses breaking news, sports, entertainment, and commercial stock, is unrivaled by any other picture provider. In an AI context, depth is essential. Search tools are increasingly combining text and imagery into unified responses; the use of reputable, rights-cleared visuals reduces legal risk and enhances user understanding, particularly on time-sensitive queries.
Legal context and disputes surrounding AI and images
Getty is known for its robust perspective on AI rights. In 2023, the corporation prosecuted Stability AI for alleged misuse of its copyrighted photos. At the same time, it created an FPBR-licensed generative model for enterprise buyers and collaborated with several big tech enterprises on provenance regulations. Getty has thus taken a stand as both a litigant and a licensor, implying that AI will benefit significantly from this agreement as far as AI search integration of professional photography is concerned.
The arrangement emerges as Perplexity is frequently criticized for its dataset construction and recapitulation strategy. Multiple publishers accused the startup of publishing descriptions of their work over the prior year without giving appropriate credit. When Perplexity was accused of using a Getty photograph without approval in an AI reply—Getty holds its copyright—the questions of image rights and proper accreditation of AI responses in picture use raised major problems. More recently, Reddit sued Perplexity for supposed mass unauthorized scraping of Reddit-produced content after allegedly breaching various technical thresholds.
The underlying legal viewpoint is that Perplexity argues that unprocessed constituents of verifiable, uncopyrighted truth and data on the worldwide internet is fair use, but the judicial system has yet to come to an agreement. The courts are endeavoring to determine whether copyright principles, as well as database rights and anti-circumvention codes, may be used—or even have to be used—for additional training and compilation of factual matters in finished theories. An authorized image pipeline is a relatively token piece of validation. It will not resolve the arguments about factual information, paywall provisions, robots.txt, and the host of other determining aspects of AI preparation and question regulation.
However, it will provide Perplexity with a valid framework within which to work on pictures—research and creativity have a clear-cut legal domain, and picture generators have traditionally defended rights with vigor. Worth exploring: beyond a picture generator community fact pattern, the Getty/Perplexity case might include subsequent influence on publishers, Perplexity, and contemporary creators.
Potential impact on publishers, Perplexity, and creators
For publishers, the promise of credits and source links alongside images might help reinvigorate the referral pathway many fear is eroding, given the decline in clicks when AI does not just answer the question. Whether the links lead to meaningful traffic will be a matter of their placement and design choices, but with Getty involved, the incentives are clearer: creators get paid through licensing, and outlets get welcomed attribution at the moment of discovery.
Expect tighter integration of content credentials through other licenses as well. Getty is part of the C2PA coalition that standardizes provenance metadata for media; if Perplexity integrates those signals, users might see cryptographic signatures that confirm details such as:
- When the image was taken
- How the image was edited
- How the image was licensed
AI platforms are already in a race to sign data-source agreements. OpenAI has unparalleled agreements with major news groups and the Associated Press. Shutterstock and Adobe have naturally become many AI workflows’ secondary go-to sources for images. Already, news organizations like the Financial Times and News Corp experiment with licensed summaries and AI integrations that keep their brands and bylines. The point is clear. Licensing is becoming a key requirement for AI that wants to be both complete and compliant.
Regulators are also keeping an eye on how all this is unfolding. Key disclosure requirements in upcoming AI legislation are centered on training data transparency and content provenance. Deals that bundle rights, attribution, and auditability are likely to fare better with watchdogs—and with enterprise customers careful with IP.
Execution challenges and the path forward for AI search
The near-term test here is execution. On Perplexity’s end, that means consistent credits, clear source links, and strong guardrails against mixing licensed images with scraped or ambiguous media. On the business side, it means proving that better visuals improve engagement without just cannibalizing publisher traffic.
If it works, the Getty partnership could become a model for how AI search can blend rapid answers with professional media in a way that respects rights holders. If it falters, it will be a case in point of how difficult it is to retrofit licensing into AI products that were built on ingestion of the web at such broad scale. For now, though, Perplexity is pursuing a path the industry has been edging toward—consent, compensation, and clearer attribution that is baked into the product experience.
