Artificial intelligence (AI) has now become an integral part of daily life. The accessibility of generative AI technologies, particularly tools like ChatGPT, has already facilitated the emergence of an AI ecosystem. AI technologies are now widely used across all fields, from education to health care, transportation to infrastructure, finance to the defense industry, and service sectors. Art is one of these fields. It quickly became evident that AI has immense potential to introduce new forms of artistic expression. In particular, the growing interest in AI-generated artworks in recent times has sparked heated debates about how this trend will impact and transform the art market, involving artists, galleries, museums and collectors.
As AI technologies rapidly transform habits, productivity, processes and management methods across all fields, the risks they pose have also begun to be discussed – albeit not as extensively as their benefits. In the field of art, initial debates resembled those that arose between photographers and painters when photography was first introduced. However, these discussions have now evolved into much more complex dimensions. OpenAI’s deep learning model DALL-E, which can generate innovative images from natural language descriptions, and its newer versions have made a significant impact not only in the art world but also among individuals of all ages and skill levels in recent years. While some argue that these advancements democratize art and support its development, others criticize them for negatively affecting artists and diminishing the value of artistic creation.
Just as in other fields, AI learns from existing data in the realm of art and generates responses based on given commands after the learning process. Therefore, existing data functions as a form of memory. Therefore, another major concern in the application of AI technologies in the field of art is the copyright of the memory. AI systems, trained on textual, visual and auditory data, learn by utilizing these datasets and generate new content based on given commands. Ongoing debates revolve around how the products created by AI should be evaluated within the framework of copyright law. In this context, the initial debates were sparked by Hollywood screenwriters. When AI is used as a tool for writing scripts for films or television shows, questions arise about whether the data used to train these systems falls under copyright protection. For instance, the Writers Guild of America (WGA) emphasizes that the use of previous scripts in training AI algorithms should be evaluated within the framework of copyright law. Otherwise, writers could find themselves defenseless against a tool trained on their own works, which were used as raw material for AI development.
A similar situation recently emerged in copyright infringement lawsuits filed by major music companies such as Sony Music and Universal Music Group against AI-driven music production companies. If these lawsuits result in rulings against AI companies, beyond the financial penalties, they will likely trigger complex discussions regarding copyright and ethical violations related to products created with AI assistance. These debates are expected to intensify in the visual arts sector as well, where AI technologies are increasingly utilized.
In this context, there is another dimension that needs attention – one that is often overlooked in these debates: the artist as a human being. The process of becoming an artist is arduous and continuous. An artist’s interactions with their environment add depth to their life, while their works emerge as the outcomes of this ongoing process. In other words, art is not just about the final product; it also enriches the human experience and deepens one’s existence. What truly matters in this process is the artist’s journey of becoming – continuously evolving through experiences, expanding creativity and developing new ways of perceiving life. The tangible works are merely the results of this journey, serving as living witnesses to an artist’s struggles, learning, and growth. Through their works, we bear witness to an artist’s formation.
However, the increasing use of AI technologies in artistic creation threatens this dimension. In AI-generated works, the artist does not undergo an experiential process of creation. Instead, they are reduced to a mere operator, directing the AI through commands. In this sense, the relationship between humans and AI remains at a technical level, leading to a growing alienation from the final product.
Complex and challenging experiences push artists into different phases of their creative journey. As a result, their works serve as witnesses to their process of becoming. While AI can replicate works that resemble those produced with deep emotions, it severs the connection to the underlying experiences and emotions or freezes the artist in a particular phase. In this context, Joe Attard provides an example of AI’s relationship with art by referencing a specific period of Picasso’s work: "For instance, Pablo Picasso’s ‘blue period’ – in which his colour palette and subject matter became notably dark and sombre – was inspired in part by the suicide of his friend Carles Casagemas in 1901. An AI perceives Picasso’s art merely as an assortment of shapes, colours and defined values. It can imitate the look of these works, but cannot grasp the emotions that inspired them. You could ask an AI to make a ‘sad’ image, but even if it produced pictures in dark colours of people crying, it wouldn’t understand the content of sadness, because it has never been sad, nor happy; nor has it experienced any other feeling."
With the integration of AI technologies into creative processes, this deeply human aspect of artistic creation is being compromised. The process of producing an artwork no longer provides the artist with an instructive or developmental experience. Instead, the artist is reduced to a mere operator, directing the AI through commands. In this sense, the relationship between humans and AI remains purely technical, leading to an increasing alienation from the final product. The phases in which artists experience their environment, relationships and introspective emotional states are disrupted. Moreover, the social benefits of art are also distorted. What does an AI-generated work add to the artist who merely provides the command? Furthermore, when audiences encounter these works, whose creation are they truly engaging with, and how will they relate to them? Will they connect with the artist through the work, or will their engagement shift toward AI? If they turn to the artist, what aspect of the artist does the work represent? If they turn to AI, what meaning will they attribute to the artwork?
A recent study found that people tend to perceive artworks created entirely by AI as less valuable compared to those made by human artists. The study also indicates that artworks co-created by human artists and AI are considered more valuable than purely AI-generated works but less valuable than those created solely by human artists. In short, when people are aware that an artwork was produced by AI, their perception of its value decreases.
As generative AI continues to evolve, its capabilities will reach new levels. This raises the question, as discussed in the literature: Will AI become an artist in its own right? This debate closely resembles discussions about whether AI should be credited as an author in scientific publications. With the emergence of papers listing AI as a co-author, academic journals have had to reconsider their policies on how AI’s contributions should be acknowledged. For instance, Science, one of the most prestigious journals, has explicitly stated that not only can AI tools like ChatGPT not be credited as authors, but AI-generated text cannot be used in published articles at all. However, over time, most journals have adopted more flexible policies, maintaining that AI cannot be considered an author but that its contributions must be transparently disclosed in the publication.
Similar debates are unfolding in the art world as well. For instance, in 2022, the U.S. Copyright Office ruled that works created by AI are not eligible for copyright protection. The rejection of Stephen Thaler’s copyright application for the AI-generated artwork "A Recent Entrance to Paradise" signaled that without human involvement, a work cannot be legally registered. Just as AI cannot be recognized as an author in scientific publications, it will not be considered an artist in the realm of art. In essence, this decision highlights the issue of the artist’s alienation from their work. When alienation occurs, the resulting piece is no longer considered the artist’s creation, and consequently, it does not qualify for copyright protection.
Although people tend to perceive AI-generated works as less valuable, in the long run, considering AI’s production capacity and rapid development, these works will become much more accessible and affordable despite their lower perceived value. Consequently, automation is expected to become more dominant in the art world as well. The only viable option to resist this overwhelming and economically challenging process seems to be for human artists to collaborate with AI. However, even in this scenario, fundamental artistic qualities such as passion and patience – key aspects of an artist’s creative identity – are likely to diminish over time.
On the other hand, as AI – essentially a black box – takes on a more active role in creative processes, human artists will increasingly find themselves as outsiders, alienated from portions of the work they contribute to. In the long run, human artists will drift away from the challenging experiences that foster learning and self-improvement. The intricate and transformative relationship between artist and artwork will deteriorate, leading to a decline in artistic diversity and depth. Ultimately, societies will be deprived of the unique value that human artists bring to art. Furthermore, as AI continues to incorporate these collaboratively produced works into its training data, the AI feedback loop will become even stronger. In contrast, the presence of authentic human-generated data will gradually diminish, deepening the sense of alienation. Over time, this cycle will not only redefine artistic creation but also reshape the role of the artist in ways that could significantly weaken human agency in art.
