Creating a photograph with AI prompts.

This is one of the most debated questions in contemporary art, and it deserves a straightforward answer instead of a vague one.

The ethics of this issue depend a lot on how much is disclosed.

The main ethical question isn't whether AI-generated images count as art. It's about being honest with buyers. If someone sells an AI-generated image as a photograph, suggesting it was made with a camera, a lens, and a person in a real place, but doesn't say so, that's simply misleading. The buyer thinks they're getting something made with a certain human skill and presence, but they're not. This is a clear-cut issue.

If the work is sold openly, for example by saying, "this is an AI-generated image, here is my process," the ethical situation changes a lot. Now the buyer knows what they are getting and can decide for themselves. Some people will value it, others won't, and that's okay.

Things get less clear with composite work, like photographs that include AI-generated parts or images that use AI a lot in editing. In these cases, the rules about disclosure are still being worked out, and photographers are still debating where to draw the line.

There's another ethical issue that often gets overlooked: the problem with training data. Most big generative image models were trained on copyrighted photos taken from photographers and artists without their permission or payment. Selling work made with these models is built on that unresolved problem. This doesn't mean the images have no value as art, but it does mean the ethics are still complicated, even if everything is disclosed.

Is it already an art form?

In my opinion, it already is an art form, but it's still new and not clearly defined. The more important question is what kind of art it is and what real skill looks like in this area.

Some people quickly dismiss it by saying, "anyone can type a prompt, so it's not art," but that argument doesn't work. Anyone can buy a camera, but that doesn't mean photography isn't art. The real question is whether there is real craft and intention in the process. There can be: knowing how to write prompts that get certain results, building a consistent visual style, choosing the best outputs instead of just taking the first one, and mixing AI tools with other methods. These all take real skill and judgment, even if they're not the same as using a camera.

It's less clear whether just generating images by typing prompts, picking an output, and selling the print is a deep enough practice to support serious art over time. The best art usually comes from processes that challenge the artist, require problem-solving, hands-on work, and even failure. A tool that gives you what you want on the first try doesn't challenge you in the same way. Whether that challenge matters for the quality of the art is still an open question, but I think it does.

Where is it going?

I believe AI image generation will become more widely accepted as a real creative medium, with its own style, artists, and discussions, just like video art or digital art eventually were. The pushback from photographers and artists makes sense, but it's nothing new. Every new image-making technology has faced this kind of resistance.

But the work that will last in this medium will come from people who use it with real intention and thoughtful ideas, not just those who make images because it's easy and profitable. The problem of abundance is real. When anyone can make a striking image in seconds, just having visual skill isn't enough. What matters most is voice, vision, and meaning. These have always set art apart from decoration, and AI doesn't change that.

Photographers and artists who focus mainly on technical skills should probably pay attention to these changes. Those whose work is centered on ideas have less to worry about.