It seems like you can’t talk about anything without the subject of artificial intelligence, or AI, coming up. It’s everywhere, and it’s currently being used for so many different purposes, for better or worse. But its use also raises a lot of questions, such as can an AI be visually creative… or more importantly, should it? 

You’ve undoubtedly heard of AI art in some form or another since it’s had some pretty high-profile exposure. Take Willy’s Chocolate Experience, for example, an unlicensed Charlie and the Chocolate Factory event organized in Glasgow in 2024. The promotional work featured AI-generated images suggesting a high quality endeavor, but visitors were so disappointed by the reality of the lackluster warehouse hosting the event that the police got involved. 

The AI promo material was so realistic it fooled many into believing they were looking at photos, which is an impressive feat when you look at it objectively. However, it was misleading in this particular case, and raised a great many warning flags for how AI art might be used in the future. 

So how does a computer create art? Well, it’s not actually that different to how it’s used to write stories. That is to say, it’s trained on the work of others. In layman’s terms, a database of reference images is given to the AI which uses machine learning to analyze composition, techniques, styles, textures and any other number of things. It then uses this to create its own works, usually prompted by words or descriptions from us humans. 

Now, there’s no denying this is an incredible feat of technology, and the AI “artist,” if you want to use the term, is capable of some impressive pieces of work. It has raised some questions which result in some debate, though. The first of which is whether the databases are being used with the original artist’s consent. 

There are many artists out there who have not given permission for their work to be replicated. It’s a complicated situation, because currently there’s no law against AI using these pieces as reference, and that’s where the waters are muddied. Because when you think about it, human artists train the same way: they study other works, add them to their own database of references, and replicate them, changing or combining them to create new works. So does the fact that it’s a machine using these references change things at all? That’s down to the conscience of the individual. 

There is such a thing as ethical AI, which is to say machine learning that only uses databases artists have consented to. In this case, there’s no ethical dilemma. One can enjoy the beauty of the resulting work and marvel at the technological achievement. Unsurprisingly though there’s always two sides to the same coin. 

Perhaps what could be called a middle distance are those who use AI art as a base. You see, much like humans, computers have yet to perfect their craft and they do still make mistakes. Sometimes there is blurring or pixelation, other times there are strange color gradients or mistakes in limbs and digits. Some artists generate an image and then create over the top of it, correcting any of these mistakes to combine the speed of a computer with the discerning eye of an experienced human artist. 

However, there is an argument to be made for AI art negatively impacting the career of human artists. As previously mentioned, a machine can learn faster and produce pieces much quicker than a human can and it doesn’t need to be paid for its services. This has drawn a lot of attention from big companies who want to save money by eschewing human artists in favor of machines. The result? A loss of jobs, particularly in the field of concept art and those who design assets for video games. That’s not to say there’s no place for such things. Not by a long shot.

After all, there are lots of small companies or solo creators for video games and the like who don’t have any artistic skill themselves and can’t afford to employ artists in the first place. There’s no doubt that AI art can help them achieve their dreams which may otherwise be unattainable. 

Another point of contention are the AI artists who sell their creations. Many sources discourage this. Because AI art is still in its infancy, the laws and regulations behind it have yet to be thoroughly implemented but tech websites such as MakeUseOf believe that using AI for profit is unethical. Since the artist who uses AI generated images doesn’t legally own the image itself, the argument is that you shouldn’t be earning money for it. 

Art being available to everyone – even those without the technical skill – is not a bad thing, though. What about the people with disabilities, who wouldn’t be able to create art otherwise? There are so many conditions out there that make art inaccessible for some. Neurological disorders, disabilities affecting the hands or concentration, even blindness – all these things can shatter the dreams of artists who want to share their creativity with others. But AI art can allow them to express themselves, which is not only a life goal for some, but a form of therapy for those who wouldn’t be able to partake otherwise. 

Whichever way you look at it, AI is not a good or a bad thing in and of itself. It’s a tool that people can use for good or ill, depending on their implementation. So the real question isn’t whether AI art is bad, but rather whether individuals are using it correctly, and as with any artist ethical practices will define your intentions.