What is AI-generated art? How is it changing the world of contemporary art? These are just a few of the questions that arise whenever we encounter artworks created through Artificial Intelligence. Amidst controversies and future opportunities, this new technology represents one of the most significant revolutions in the creative field in recent years. However, it also carries a dark side that should not be underestimated.

Farming, l’opera di Federica Di Pietrantonio in mostra dal 24 maggio a Roma negli spazi di WEGIL | Courtesy Federica Di Pietrantonio
Let’s start from the beginning. AI art allows for the creation of original images based on a prompt, which is a textual instruction or command. Platforms like Stable Diffusion or Midjourney make this possible by utilizing neural network technology. Prompts can provide instructions of any kind, significantly influencing the final image outcome. In short, AI generators are trained using a vast number of images, often sourced from the internet, to generate new illustrations without prior references.

Artisti che lavorano con l’Intelligenza Artificiale
There are numerous artists working in the field of AI art, offering ample opportunities for contemporary art. One notable example is Refik Anadol, who was among the first to work with this technology during an artistic residency at Google seven years ago. His artworks transform and react to the viewer’s movements or external agents, never repeating themselves. “Unsupervised,” Anadol’s artwork exhibited at MoMA, reflects on technology, creativity, and modern art.
Another artist working in this field is Lawrence Lek, whose research focuses on exploring the legal opportunities of Artificial Intelligence as it manifests itself both in pop culture and in the broader sense of technology.

Black Cloud (2021), un’installazione di Lawrence Lek | Courtesy Lawrence Lek
In Italy, last summer, Roberto Fassone attempted to metaphorically feed an Artificial Intelligence with hallucinogenic mushrooms to obtain outputs capable of revealing some form of unexpected consciousness at Combo, in Turin. Mara Oscar Cassiani peculates about the possibility of falling in love with Artificial Intelligence. “Ai Love, Ghosts and Uncanny Valleys <3. I broke up with my Ai and will never download them again” is her artwork included in Spark and Frictions,a group exhibition dedicated to the relationship between contemporary art and AI, held from May 24th in Rome at WEGIL, a cultural hub of the Lazio region. The exhibition features Italian and international artists, including Federica Di Pietrantonio, Alice Bucknell, Yue Huang, and many others.

Ai Love, Ghosts and Uncanny Valleys <3. I broke up with my Ai and will never download them again | Courtesy Mara Oscar Cassiani
The dark side of AI art
The dark side of AI art arises from the exploration of new possibilities and the opportunities for collaboration among various mediums and artists. However, this also gives rise to various ethical issues. The use of such extensive databases raises significant copyright concerns. One prominent case involves a group of artists – Sarah Andersen, Kelly McKernan, and Karla Ortiz – who have filed a class-action lawsuit against some companies behind AI generators, claiming that these services violate copyright laws and engage in unfair competition. Specifically, the class-action targets Midjourney Inc, DeviantArt Inc – responsible for DreamUp – and Stability A.I. Ltd, the company behind Stable Diffusion. Getty Images has also joined the lawsuit, accusing Stability A.I. of finding over 15,000 photographs from its archives in the Stable Diffusion dataset. It is still unclear whether a precedent will be set, potentially leading to a monopoly over these images, or if an agreement will be reached. However, there is a new development. Ascendant Art is a new AI-generated art platform that promises to commit to paying royalties to all artists who submit their works to train the algorithm. Will it succeed?

Courtesy Ascendant Art
The dark side of AI art remains tied to everything that Artificial Intelligence has absorbed and can apply. Without having yet defined the boundaries of this technology, we can only wait to see what legislative measures will be implemented to protect artists and viewers.