Getty Images unveils AI-powered art tool to generate custom images

Getty Images has unveiled a new AI-powered art tool called "Generative AI by Getty Images," developed using an AI model from Nvidia and trained on a part of Getty's library, which includes around 477 million stock content assets, and can generate images based on textual prompts like other Generative AI LLM such as DALL-E 3 or Midjourney

Upon creating and downloading visuals, users will receive Getty's standard royalty-free license, which offers protection against copyright lawsuits and allows for unlimited worldwide nonexclusive use. The tool has safeguards to prevent misuse for disinformation or replicating the styles of living artists. All images produced will carry a watermark indicating they are AI-generated. This sounds quite similar to what Shutterstock did earlier this year with its own version of ethical AI generation.

The tool is accessible via Getty's website or can be integrated into apps and websites through an API. It also provides customization options for users to create images in line with a specific brand style or design language. While the pricing will be separate from standard Getty Images subscriptions, it will be based on prompt volume. Getty retains the right to retrain its model using these images, and contributors whose works are used in training the AI model will be compensated, so the revenue generated from the tool will also be shared with these contributors. The launch of this tool marks Getty's entry into the generative AI sector, a field that has already been widely embraced by other major alternatives such as the previously mentioned Shutterstock or Adobe Firefly, whose direct integration with tools from the Adobe Creative Cloud suite like Adobe Photoshop or Adobe Illustrator gives it a significant advantage.