There are many AI image generators. But they often have one problem: copyright. First, it is not always clear where the training material comes from. Second, it is not certain to what extent you can get into legal trouble with the generated images.
Such image generators are often out of the question for companies and other organizations because they are (or seem to be) too risky to implement.
This is where an opportunity lies for vendors whose business is directly threatened by AI image generators: Stock photo platforms.
I have now been able to test “Generative AI by iStock”, a collaboration between image giant Getty Images and hardware specialist Nvidia. It promises “built-in legal protection”. I received credits for 100 generated images.
To judge the quality of the generator in detail, I would have to test it longer and more intensively. My impression is that it is on the level to be currently expected. In other words, good results can be achieved – sometimes with more, sometimes with less effort. Typical problems such as deformed hands are also present at times.
Useful functions include “Refine” to change certain parts of an image, and “Extend” to create a wide format result from a square template, for example. These functions are also available for images from the existing iStock catalog. I think this is a good idea because I often find that an image almost fits what I’m looking for and need, but I don’t have the resources to adjust it.
I liked the interface of the image generator. It makes it easy to select the image format or a color mood. A helpful feature is the ability to use an already fitting image as a starting point for other variations with just one click.
As usual, everything starts with a prompt. With iStock, there are input fields for the main subject, an action, the environment, and visual aesthetics. This ensures that the desired result is described with enough precision. A negative prompt is also available to avoid certain elements in the results.
However, it is not always easy for non-experts to fill in these fields. When I go looking for stock photos, I often have only a vague idea of what I want. Instead, I take a step-by-step approach to the end result and draw inspiration from what I find on the platform along the way.
When asked about this, Grant Farhall, Chief Product Officer at Getty Images and iStock, explained that they have introduced a “lookbook” for this purpose. It collects examples of motifs and styles that can serve as inspiration.
In May, companies will also be able to define general specifications to align the images generated more with an internal style guide.
Bottom line
Even in this early version, iStock’s image generator is an interesting offering for businesses. It is a useful addition to the stockphoto platform. A video AI has also been announced – in collaboration with the startup Runway.