%20(3).png)
Have you ever tried creating your own AI-generated avatar on TikTok or turning your selfie into a stunning piece of art? It's an incredibly fun and exciting experience! AI art has become a captivating trend that is flooding social media platforms. People now willingly pay fees to access image generation tools, while others take on the role of an AI artist, generating images and even profiting from their creations. But is AI art just a passing trend, doomed to fade away like so many others? Or does it have the potential to usher in a new era of artistic expression, much like photography once did? The question remains, however: will AI art be universally accepted as a legitimate new form of artistic expression? As the media spotlights this emerging trend, it also becomes clear that there are two distinct groups of people with opposing views. On one side are those who enthusiastically embrace this new art form and collaborate with AI to create groundbreaking and innovative works. On the other side, there are those who vehemently oppose this trend and consider it unethical or even illegal. The ethical and social issues surrounding AI art have become a focal point of discussion, polarizing opinions and sparking important debates about the nature and validity of AI-generated artworks. Is there a copyright for this kind of creation? Who is the owner? Has the data behind this AI been lawfully collected? Will AI art reflect the biases of its creators and perpetuate existing inequalities?
Before we get into the main issues, let's look at when and how these systems came to be. AI art isn't really a new concept. In the 1950s, Alan Turing, one of the "founding fathers" of artificial intelligence, had the idea of : If humans make decisions and solve problems by using existing information and analyzing it, why can't machines do the same? So basically, it is a science whose goal is to let machines do what humans do with their intelligence. In the late 1960s, AI art first appeared with Harold Cohen, who used an AI system to create art drawings, using a symbolic, rule-based approach to generate black-and-white images that Cohen then painted. After that, AI progress has not stopped until 2022, and several AI-generating systems have been developed. Text-to-image generators in particular are currently experiencing a real boom. These systems generate art using machine learning and deep neural networks. Large datasets are used to train and teach these algorithms to find patterns and styles to create new art. The datasets used are collected and scrapped from many sources on the Internet.
Although this new technology is a great success, some artists complain about it for many reasons, and there are even communities that do not support AI art.
They imply that the works created by machines are not considered art. They use terms like "steal," "plagiarize," and "scrap" to describe the way machines create artwork. This is an overly reductive and inaccurate description of the way Machine Learning models create images. The algorithm learns a particular aesthetic by analyzing thousands of images, and then attempts to generate new images that match the learned aesthetic. So what is generated is considered art? Let us look at Wikipedia's definition of art: "Art is a diverse range of human activities and resulting products that involve creative or imaginative talent expressing technical skill, beauty, emotional power, or conceptual ideas." Based on this definition, AI-generated images are indeed art in the true sense. But the difference between human art and AI art lies not in the nature of the product, but in its creator. So maybe we should define the word "artist" instead of defining "art." A human who creates art is considered an artist. Is the machine also an artist? Or is it the creator of the machine? Or is it the end user of the machine? These questions cannot be answered because there is no specific law that defines the ownership of these creations.
As AI technology advances and AI-generated art gains recognition and commercial value, there is a growing need to review the copyright framework and determine how it can accommodate and protect the rights of AI-generated artwork and its creators.
To benefit from copyright protection, the work must be original. A work is original if it reflects the personality of the author, that is, if it represents a personal and intellectual contribution. This is a subjective concept that only a judge can definitively decide. To meet this criterion of originality, the author can usually only be a human being. The word “original” is confusing. In the ordinary sense, that which is not commonplace is original. In copyright, the original is that which springs from the human mind. Once a work is original and formalized, it is protected by copyright, regardless of: its genre (graphic, literary, musical, choreographic…), its form of expression (visual, on a material support or oral), its merit and its purpose. AI-generated art, then, cannot be protected by copyright because the creative process lacks the human effort, skill, and judgement to create originality. But the legal framework for AI-generated art and copyright is likely to evolve. Policymakers and legal experts are actively addressing these issues to develop guidelines and regulations that balance recognition of the contributions of human creators with consideration of the unique characteristics of AI-generated works.
Since there is no copyright for art created by AI. So there is no problem with commercialization or ownership of such creations. However, the real problem starts at the first stage of creation: collecting data. AI systems cannot function without data. The more we feed them data, the more they learn. For example, millions of pieces of data are scraped from the Internet, and most of these data are copyrighted images that the owner never gave permission to use. Some other data was available on the Internet for research purposes, not commercial applications. Some artists are taking action against the companies that develop these systems. They claim they never gave permission to use their artwork to train the models. Some others complain that their name is included in the prompts so that the user can create an image inspired by their art style. Many artists have indirectly contributed to the creation of these artworks but have not benefited from them. But the user who created the image and wrote the prompt is the only one who can benefit, and sometimes there are companies that charge royalties. So artists see AI art as a threat because it takes jobs away from human artists, designers and illustrators by using their work without their permission. So we may need to establish a policy for the use of these generating systems.
As AI systems are trained on large datasets, there is a possibility that they will reproduce biases present in the data, leading to the generation of biased content that reinforces negative stereotypes and perpetuates discrimination. For example if we ask an image generator to generate a picture of a nurse it will generate women and if we ask it to generate a picture of an engineer it will generate men. Another problem is the lack of transparency and explainability of generative AI biases. AI algorithms can be very complex, making it difficult to understand how and why biases occur in the generated content. This lack of transparency can hinder efforts to identify, eliminate, and effectively mitigate biases. It also raises questions about accountability and responsibility for the biases present in AI-generated art.
The emergence of AI-generated art has sparked both fascination and controversy. It has opened new avenues for creativity and challenged traditional notions of authorship and artistic expression. As the field of artificial intelligence evolves, so will the ethical considerations surrounding artificial art. It is our collective responsibility to remain vigilant, to adapt to new challenges, and to foster a future in which AI and art coexist ethically, inspiring creativity and enriching our understanding of the human experience.