Getty Images' Leap to AI-Powered Image Generation

• Generative AI

As artificial intelligence continues to evolve and reshape multiple industries rapidly, Getty Images has stepped into the spotlight with its recent launch of an AI-powered image generator.

Getty Images, a renowned player in the supply of stock images, editorial photos, videos, and music, has unveiled an AI art tool named Generative AI by Getty Images. The tool was developed in technical collaboration with Nvidia and claimed to be "commercially safer" than similar solutions. Generative AI by Getty Images was trained using a significant portion of Getty's extensive library, boasting approximately 477 million assets.

Generative AI by Getty Images

Generative AI by Getty Images works like popular text-to-image platforms like OpenAI’s DALL-E 3 and Midjourney. It renders images based on text descriptions or prompts. For instance, a prompt such as "photo of a sandy tropical island filled with palm trees" would be transformed into a visual representation by this tool.

Customers who use this tool will receive Getty’s standard royalty-free license. This license includes indemnification, protection against copyright lawsuits, and the right to "perpetual, worldwide, nonexclusive" use across all media. However, it’s important to note that the use of the tool is not entirely unrestricted. Getty has put in place safeguards to prevent misuse of the tool for disinformation, misinformation, or replicating the style of a living artist.

The tool, for example, will not allow a user to create an image of public figures such as Joe Biden in front of the White House or a cat in the style of Andy Warhol. Further, all impressions generated by the tool will carry a watermark identifying them as AI-generated.

Responsible AI Tool

The CEO of Getty Images, Craig Peters, has emphasized the importance of developing a responsible tool that customers can trust for commercial use. He also confirmed that any content generated by the tool will not be added to Getty's content library for others to license. However, Getty reserves the right to use the images to retrain the model. Additionally, Getty plans to compensate contributors whose work has been used to train the AI model and share the revenues generated from the tool.

The tool can be integrated into apps and websites via an API or enabled on Getty’s website. Customers can customize the device with proprietary data to create images that align with a specific brand style or design language. The pricing for this tool will be separate from a standard Getty Images subscription and will be based on prompt volume.

According to Grant Farhall, chief product officer at Getty, the service allows brands and marketers to safely embrace AI and expand their creative possibilities while compensating creators for including their visuals in the underlying training sets.

Legal Landscape in AI

Before this launch, Getty was a vocal critic of generative AI products like Stable Diffusion, trained on a subset of its image content library. Getty had even sued AI startup Stability AI, involved with creating Stable Diffusion, for allegedly copying and processing millions of its images and associated metadata without informing or compensating Getty contributors.

Getty is not the only company exploring safer and more ethical approaches to generative AI. Other companies like Bria, Ascendant Art, Shutterstock, and Adobe have also been developing similar tools and compensation models for contributors. As the AI landscape evolves, companies must adopt ethical practices that respect intellectual property rights and appropriately compensate contributors.