As my understanding of media literacy continues to deepen through my engagement with this online community; Generative AI begins to present itself more and more. Many are concerned, including myself to an extent, that AI-derived artwork created by models trained on copyrighted material will flood the market and weaken copyright holders style. Understanding the concerns, I sought to find two things:

1. Is the use of Generative AI to create art infringement of any specific artist and, if so, what warrants willful intent?
2. What kind of compensation is just, if any, for the unauthorized use of any specific artist’ work?

The above photo is a stock photo of Rainbow Bridge I found on Creative Commons Search Portal to present another view different from my own. Both of these photographs exist online under creative commons and could possibly be used to train an AI model. Does the availability of the images online mean that anyone can use it? In some cases yes, but in a majority the answer is a hard maybe. It’s important to understand how the internet works to understand the complications faced with limiting access to publicly-shared information, avoiding the way-back machine, and the many other questionable means of online information being propagated and replicating. Copyright has been a problem online before Generative AI existed, and though many models were trained on massive database of information much of this information was scraped from the web itself likely including copyrighted works.

The defense states that AI creates transformative content that pulls from many creative sources, so it can’t take an artist’ style; this has been challenged in a class action suit as artist come forward with examples and data showing potential infringements. The logs showed how many users searched for a particular artist style and would generate similar content; would this count as willful infringement? In my opinion, the people who train the AI model should compensate the most duplicated artist when they’re aware they’re being heavily imitated for commercial use, perhaps even limit creations based on copyrighted works from being downloaded or shared if the company hasn’t obtained the artist’ work through legal means.

To test Generative AI for myself, I decided to use the following prompt to generate the photo above:

An upward view of the Rainbow Bridge in Tokyo, Japan; no people mid day

The platform had an option to enhance the prompt and I decided to give that a try, it used the following prompt:

Create a captivating image featuring an upward view of the iconic Rainbow Bridge in Tokyo, Japan, bathed in the bright light of midday. The bridge towers majestically against a vivid blue sky, its colorful arches reflecting sunlight. Wispy, cotton-like clouds drift lazily, adding depth to the serene atmosphere. The surface of the water below glitters like shattered glass, mirroring the vibrant hues of the bridge. Lush greenery lines the edges of the scene, with occasional bursts of blooming flowers adding splashes of color. For an artistic style, opt for a photorealistic rendering that captures the intricate details of the bridge’s structure and the serene beauty of the surroundings, making the viewer feel as if they are standing right beneath this architectural marvel.

The enhanced feature really did some heavy lifting! I was surprised at how detailed the enhancement was and it really made the image choice quite easy. Looking at the photo above after my initial description was impressive, but what would have been the image options had I not used the enhanced option?

I’m thinking we might have some time to work out a structure that can support professional artist whose work was inadvertently used to create these tools; however, I believe the examples above show the difference the prompt makes in the image’ outcome. There is little we can do to change what’s been done to create these tools, it’s up to us now to find ways to use them ethically and justly; that means it is vital we use attribution in the content we share as users and creators. Learning how Creative Commons works is an integral part of contributing to the global community and that means taking the time to properly attribute any media included in your content, including your own. I am taking everything I’ve been learning in my classes and locking it in by sharing it here in my public digital journal; Thanks for reading and I’ll write again soon!