Earlier this year, stock photo service provider Getty Images sued Stability AI over what Getty said was the misuse of more than 12 million Getty photos in training Stability’s AI photo generation tool, Stable Diffusion.
Now Getty Images is releasing its own AI photo generation tool, which will be available to its commercial clients. And for that the big dog is called in: Nvidia.
The tool, simply called Generative AI by Getty Images, is available on the Getty.com website. It will also be available via an API, so Getty customers can connect it to other apps. It is designed strictly for commercial use; For example, a photo editor or marketer would need to find a generic image of a sneaker or a smartphone, and instead of using a standard stock photo, they could ask the tool to generate something new. (Getty has said she doesn’t expect news organizations to use it.)
The Getty AI tool is trained entirely on Getty Images – hundreds of millions of them – and uses Nvidia’s model architecture, Edify. Craig Peters, CEO of Getty Images, says that thanks to the company’s partnership with Nvidia, we have “virtually unlimited [graphics processing units], something almost no one has these days. We could go through countless training sessions with Nvidia and their team to get this right.”
Peters says Getty does not pay for access to Nvidia’s technology, nor does Nvidia pay Getty for its content. “We are partners. Just partners.”
By launching this tool, Getty is competing with rival Shutterstock, which is working with OpenAI to allow the latter to train its Dall-E models on Shutterstock images; as well as Adobe, which recently put its generative AI engine, Firefly, in Photoshop. Adobe Firefly is trained on “hundreds of millions of high-resolution Adobe Stock images, openly licensed content, and copyright-expired public domain content,” according to Adobe’s website.
Artistic license
Getty’s dip into the AI photo pool raises questions about the ethics of training an AI model based on more than two decades of photographers’ images and how companies exploring this business model will ultimately pay these photographers.
In terms of licensing, Peters emphasizes that Getty’s AI photo generator is different from other AI image tools because Getty has acquired the legal rights to the photos used to train the models. “It’s commercially clean and that gives us the opportunity to back it 100 percent,” he says. “So if you want to use generative AI to be more creative and explore new boundaries, Getty Images is the only offering that is fully reimbursed.” That means that if a customer downloads and uses an AI-generated Getty image, and a Getty contributor points out that it closely resembles their original artwork, Getty promises the customer that it will be covered by Getty’s royalty-free licensing agreements.
Getty also said in a press release that customers using the tool will have the right to perpetual, worldwide, non-exclusive use of the images and that new AI-generated content will not be added to Getty’s existing content libraries for others to use .