Ethics of AI in Art: Bias, Originality, and Consent

The deflection of the data can draw out biased results when humans experience social discrimination or prejudice. To solve this, we have to build up learning data that cover various human groups such as age, gender, and race, and revise and evaluate fairness throughout the development of the AI model. Because of the unauthorized use of AI data, many AI models collect works published on the web without permission from rights holders and use them for learning. It leads to continuing controversy related to copyright infringement and unfair competition prevention laws, and it results in legal disputes with media companies and individual creators. To address this, it is important to purchase a legitimate, copyright-risk-free dataset or secure it through a licensing agreement. Also, we need a clear legal interpretation of the principle of fair use and an explicit disclaimer against the use of a work for learning. The legal right to ensure transparency in data use and to seek suspension and damages in the event of an infringement of rights must be guaranteed.

Whether the data used for AI learning is original high-quality data or data containing biased or inaccurate information directly affects AI performance and reliability. Complex dependencies between datasets might pose a risk to the legality of the entire system. Also, we have to secure transparency and manage and make notes about the process and source of data collection.

These days, I have become alert to data bias, unauthorized use, and originality issues related to AI because students usually use AI when they do an assignment. Once, I did a task with AI’s help, but I got a zero score because the content was similar to others. Also, another friend told me that she always uses ChatGPT so that she can’t write an essay herself. I think AI gives exact technology and knowledge, but if we use it incorrectly, it leads to terrible results. Because of this example, I felt that AI cannot replace human thinking and expression skills, and AI's excessive help loses learning values. AI is an efficient and fast technology, but it leads to social problems if the data are biased or are copyrighted material. Thus, when we get AI’s help, we have to critically judge and verify rather than unconditionally believe the result. Also, I think it is important to use AI as a tool but not rely on it, to maintain creative thinking and self-expression.

In order for positive development of AI, I felt that technical ethics and a sense of individual responsibility must be developed together. I realized that there are problems such as deflection, unauthorized use of AI, and originality issues behind the comfort. Especially, unauthorized use and deflection issues are problems of social responsibility, not mere technological errors. I strongly believe that our society values human creativity and judgment. AI is used for artworks and learning, but it is important to recognize ethical risks. Eventually, if we want to utilize technology, we need critical judgment and a responsible attitude.