Getty Images has introduced a generative artificial intelligence (AI) image service with the claim that it is “safe” to use because it is trained on Getty’s licensed content library and, therefore, does not have the same risk of copyright infringement as other generative programs. This new program comes after Getty’s reveal of a generative AI technology in September, introduced in a demo, and is now available on their iStock site. Developed in collaboration with AI chip giant Nvidia, the program is presented in the midst of a copyright infringement controversy, with companies and scholars contesting over copyright issues related to AI. Getty’s program facilitates indemnification for its users, limiting liability and offering further indemnification options. With a cheaper introductory offer, the initial 100 images are not as aesthetically or narratively interesting as other AI image creation programs like DALL-E and Clipdrop. However, the iStock service offers a workable alternative and is positioned as a safer option. Users can create an account by signing up on istockphoto.com and paying for a bucket of 100 image generations. The results are fairly literal and dull, and haven’t shown much creative potential. The program has difficulty parsing complex ideas or producing imaginative outcomes for fantastical prompts. Additionally, there are limitations to the program’s indemnification, as some prompts containing copyrighted or trademarked material may be blocked or fail to generate an image. Despite its indemnification efforts, iStock has potential limitations regarding protected materials.
