Getty Images, the renowned provider of stock imagery for digital and print content, has introduced a dedicated image generation service for its subsidiary iStock. The new service, called Generative AI by iStock, offers designers and business users access to a text-to-image generator that creates visually appealing images from simple text prompts. Notably, the images generated by this tool come with legal protections and usage rights, setting it apart from other image generators. This means that businesses can confidently use the generated images for commercial purposes without the fear of them being stolen for unauthorized use elsewhere.
This latest move by Getty Images in the field of generative AI follows the launch of an image generator for its main platform in September. It also comes in response to rival stock image service Shutterstock’s introduction of an AI image generator feature that adapts images from its asset library.
Generative AI by iStock: Features and Benefits
Unveiled at CES 2024 in Las Vegas, Generative AI by iStock bears striking resemblance to Getty’s image generator for its own site. Users can access a text-to-image tool, where they can describe their desired visuals, and the underlying model will generate high-quality images in resolutions up to 4K.
Getty, with its vast catalog of licensed images, emphasizes that this tool has been trained on its proprietary creative library, encompassing exclusive photos, illustrations, and videos on both Getty and iStock. This makes it an ideal choice for acquiring licensed AI content, especially at a time when artists raise concerns about unauthorized usage of their artworks for training AI models.
To ensure content quality and protection against theft, each image generation produces four distinct images, each accompanied by legal coverage of up to $10,000. Furthermore, iStock refrains from adding the generated content to its creative library for others to license, reinforcing its commitment to customer exclusivity and protection.
A notable feature of Generative AI by iStock is its prompt builder, which allows users to define various aspects of their input. Users can specify the subject, describe its actions, and provide details about the surroundings. Additionally, they have the option to define the color, mood, and aspect ratio of the desired image.
The image generation process is powered by Nvidia’s Picasso AI model library. Getty leveraged Picasso’s Edify foundation model for visual design and combined it with its licensed content to create a customized model that drives the Generative AI by iStock service.
Additional Capabilities and Integration
Alongside the introduction of Generative AI, Getty Images also announced advanced editing capabilities for its users, including inpainting and outpainting features. Inpainting enables users to mask a specific region of an image and replace it with an object defined through text input, while outpainting allows users to expand the image to fit various aspect ratios by filling in additional content based on the original image. These features will be available on both iStock and Getty Images.
Developers can integrate these capabilities into their creative applications by accessing the provided APIs immediately.
Getty’s AI tools, including Generative AI by iStock, distinguish themselves from competitors in the AI-driven content generation industry, such as Midjourney, DeviantArt, Runway, and Stability AI. Adobe has also joined the space with its Firefly model, trained on its proprietary catalog of licensed images, which is integrated into several Creative Cloud products.
