Historical Context: A Dispute Paves the Way for Innovation
Earlier in the year, Getty Images was embroiled in a legal tussle with Stability AI over the alleged misuse of a staggering 12 million Getty photos. This set the stage for Getty Images not just to protect its vast archive but to leverage it in an innovative manner. Thus, was born “Generative AI by Getty Images.”
The Partnership: Getty Images and Nvidia
Powering the tool is Nvidia’s state-of-the-art model architecture, Edify. Getty Images, leveraging its partnership, had vast access to graphic processing units, enabling numerous training runs to perfect the generative AI model. This collaboration isn’t a typical commercial exchange. Both giants are coming together as partners, signaling a new era of corporate collaboration in the age of AI.
This development places Getty in direct competition with the likes of Shutterstock, which recently collaborated with OpenAI, and Adobe with its Firefly generative engine integrated into Photoshop.
A New Age of Licensing
The significant differentiator Getty is betting on is its vast licensed image archive. Craig Peters, CEO at Getty Images, emphasized the commercial safety of their AI-generated images. The company is confident in the legitimacy of its offerings and stands behind them, promising customers the security of Getty’s royalty-free licensing agreements.
This approach addresses a significant concern in the AI photo generation world: the ethical implications of training models on decades’ worth of photographers’ works. With Getty’s proposition the rights of the original photographers are not lost in the ether.
However, this venture isn’t without challenges. Determining which photos from the enormous training set inspired a newly generated image poses a formidable problem. Getty’s approach to this is a two-pronged evaluation – the volume of content a contributor has in the training set and the performance of that content over time.
Blurring the Copyright Lines
Generative AI has brought about a paradigm shift in copyright considerations. Its nature involves ingesting vast amounts of data, often blurring the lines between training and the commercialization of new creations.
Concerns around the ethics of such technologies are evident in suits filed against tech giants like Meta and OpenAI, accused of indiscriminately scraping internet content to train their AI models. Getty’s approach to using solely its licensed images might seem like a solution, but it still presents a multitude of questions.
Looking Ahead
Generative AI by Getty Images represents a watershed moment in the stock photo industry. As Getty Images combines its vast licensed image repository with Nvidia’s tech prowess, the potential to reshape the industry is undeniable.
Yet, as with all new technological forays, the path is fraught with ethical and legal quandaries. The true success of this endeavor will not merely be its technical proficiency but its ability to harmoniously blend innovation with the sanctity of artists’ rights. Only time will tell if Getty Images has truly cracked the code.