Perplexity Signs Getty Images Deal After Plagiarism Accusations

2 days ago7 min read

The recent licensing agreement between AI search startup Perplexity and stock photography behemoth Getty Images represents a fascinating and pivotal moment in the ongoing evolution of artificial intelligence's relationship with intellectual property, effectively legitimizing what were previously contentious scraping practices under the respectable umbrella of a commercial partnership. This development arrives directly on the heels of a turbulent period for Perplexity, which last year faced a barrage of plagiarism accusations from major news organizations like Forbes and Wired; these publications meticulously documented how the AI service was not only summarizing their paywalled investigative reporting nearly verbatim but also failing to provide adequate attribution, a practice that echoes the early, Wild West days of web scraping that eventually necessitated the robots.txt protocol. For those of us who have followed the trajectory of large language models (LLMs) from academic curiosities to commercial powerhouses, this sequence of events—accusation, controversy, and then a strategic licensing deal—feels like a familiar playbook, reminiscent of the paths taken by AI giants like OpenAI when they secured content partnerships with publishers like The Associated Press and Axel Springer after training their models on vast, unlicensed corpora.The core technical issue at play here is the fundamental nature of how generative AI models are trained: they ingest trillions of tokens of text and associated media to develop a statistical understanding of language and concept association, a process that, while transformative, has always operated in a legal gray area concerning copyright, especially for multimodal models that require paired image-text data. Getty, a company that has been notoriously aggressive in protecting its vast library, having previously sued Stability AI for allegedly copying millions of its images to train Stable Diffusion models, now appears to be adopting a pragmatic, if not slightly cynical, strategy—if you can't beat them, license to them, thereby converting a potential legal adversary into a revenue stream.This deal doesn't just provide Perplexity with a clean, legal source of high-quality visual data; it serves as a powerful signal to the market and potential investors that the startup is serious about navigating the complex IP landscape, a critical concern for any AI company hoping to achieve long-term viability and avoid the existential threat of billion-dollar copyright infringement lawsuits. However, this corporate détente raises profound questions for the broader ecosystem: does this model of post-hoc licensing create a two-tiered system where well-funded AI firms can purchase legitimacy while smaller open-source projects are left legally vulnerable? Furthermore, what does 'fair use' mean in this new paradigm if any commercial application ultimately requires a licensing fee? We must also consider the perspective of content creators and photojournalists whose work constitutes the value of Getty's library; will they see a fair share of this new revenue, or are they merely pawns in a larger negotiation between tech titans? The implications extend beyond mere stock photos, touching the very foundation of how knowledge and creativity are monetized in the AI age.As we look toward an future of artificial general intelligence (AGI), the need for a sustainable and equitable framework for training data acquisition is paramount; while partnerships like Perplexity-Getty are a temporary patch, they are not a permanent solution. The industry desperately needs clearer guidelines from legislators and perhaps even novel, blockchain-based provenance systems to ensure that the original creators of the data that fuels our intelligent machines are recognized and compensated, preventing a future where AI innovation is built on a foundation of uncredited labor and appropriated artistry.

#Perplexity

#Getty Images

#licensing deal

#copyright

#plagiarism

#AI ethics

#featured

Stay Informed. Act Smarter.

Get weekly highlights, major headlines, and expert insights — then put your knowledge to work in our live prediction markets.