AI search startup Perplexity has inked a multi-year licensing deal with Getty Images that legitimizes its use of stock photos and marks a strategic pivot toward formal content partnerships. The agreement comes after months of plagiarism accusations from major news outlets and signals the company's effort to move beyond scraping controversies that have dogged its rapid growth.
Perplexity just made its biggest move yet to clean up its copyright mess. The AI search startup's multi-year licensing deal with Getty Images doesn't just give it permission to display stock photos - it retroactively legitimizes months of questionable image usage that sparked industry-wide backlash.
The timing couldn't be more critical. Perplexity has been fighting a two-front war against plagiarism accusations and legal challenges, with Reddit's October lawsuit alleging "industrial-scale, unlawful" scraping still pending. This Getty Images partnership represents the company's first major content licensing victory.
According to sources familiar with the deal who spoke to TechCrunch, Getty Images has quietly been part of Perplexity's Publishers' Program for over a year, sharing ad revenue when their content surfaces in search queries. Today's announcement formalizes a much broader relationship.
The deal structure breaks from traditional licensing models. Since Perplexity doesn't train its own foundational models like OpenAI or Google, it's not paying the typical lump-sum fees that AI companies negotiate for training data. Instead, sources suggest it's a performance-based arrangement tied to actual image usage and attribution.
"Attribution and accuracy are fundamental to how people should understand the world in an age of AI," Jessica Chan, head of content and publisher partnerships at Perplexity, said in a statement. The emphasis on attribution isn't just PR speak - it's central to Perplexity's legal defense strategy.
The startup has been arguing that its use of publisher content constitutes "fair use" because publicly available facts aren't copyrightable. But that defense got harder to maintain when outlets started documenting specific cases where Perplexity lifted entire Getty Images photos alongside scraped Wall Street Journal articles.
Nick Unsworth, vice president of strategic development at Getty Images, framed the partnership as acknowledging "the importance of properly attributed consent and its value in enhancing AI-powered products." That language suggests Getty Images views this as a template for future AI licensing deals.
The broader implications extend beyond Perplexity. OpenAI already has a data licensing agreement with Reddit, while Google has struck deals with major publishers. As AI companies face mounting legal pressure, formal licensing agreements are becoming the industry standard rather than relying on fair use defenses.
For Perplexity, this deal buys crucial legitimacy as it prepares for what industry insiders expect will be a significant funding round. The company's search product has gained traction, but copyright controversies have made some investors cautious about backing a platform built on potentially infringing practices.
The Reddit lawsuit remains Perplexity's biggest challenge, with allegations of circumventing technical barriers specifically designed to block AI scraping. But this Getty Images partnership demonstrates that major content providers are willing to work with the company when proper compensation and attribution are involved.
What's particularly interesting is how this reshapes the competitive landscape. While Google and Microsoft can leverage their existing publisher relationships and deep pockets for licensing deals, startups like Perplexity need to prove they can build sustainable partnerships that go beyond traditional web scraping.
Perplexity's Getty Images deal represents more than just a licensing agreement - it's a blueprint for how AI companies must evolve from scraping to legitimate partnerships. As legal pressure intensifies and investors demand cleaner business models, expect more startups to follow Perplexity's lead in formalizing content relationships. The real test will be whether this approach can scale across the thousands of publishers whose content powers AI search results.
