The AI’s Ethical Limitations from the Societal Perspective: An AI Algorithms’ Limitation?

This study focuses on the ethical boundaries of artificial intelligence (AI) technologies from a societal perspective. We employed a methodological approach by conducting a systematic literature review using the PRISMA guidelines. The outcome of our approach is the identification of a single work that discusses this issue from a societal perspective, highlighting the limitations imposed by the impossibility of complete modeling of ethical requirements from society, as well as the limitations imposed by the nature of the partial transposition of these requirements into AI algorithms. Our perspective regarding these limitations is based on three key concepts: (1) the categorization of domains based on varying levels of ethical significance and the establishment of a systematic approach for ethical taxonomy; (2) the necessity of constructing an ethical meta-model at the societal level, as well as domain-specific ethical models for all AI technologies; and (3) the gradual and prudent replacement of humans in the development of AI technologies, driven by their inherent limitations in this field. The originality of our approach lies in formulating a research idea related to limiting the integration of ethics into algorithms, systematizing research on this highly relevant topic, and developing an opinion about addressing ethical aspects at the level of an AI entity. Ultimately, based on our findings, we have provided a positive answer to the question “An AI Algorithms’ Limitation?”.