Ethical Agency of Artificial Intelligence

Introduction: An ethical deadlock occurs when conflicting or equally compelling moral reasons challenge decision-making. Artificial intelligence cannot, like human beings, escape from such conflicts through reflective and self-aware processes. Three fundamental conditions of moral agency - free will, awareness, and the ability to distinguish right from wrong - are conditions that current artificial intelligences lack. At best, current AI systems exhibit norm-conforming and rule-based behavior that appears ethical, but they lack the aforementioned three characteristics. Therefore, this article examines the fundamental challenge of artificial intelligence in exercising ethical agency when confronted with ethical deadlocks. Material and Methods: In the present study, a literature review was used as a research methodology. Articles published in the field of ethics and artificial intelligence were examined. Conclusion: The findings of the present review distinguish qualitatively between “functional moral agency” (behavior that conforms externally to moral norms) and “authentic moral agency” (possessing internal understanding, autonomy, a sense of responsibility, and entitlement to blame and praise), and argue that artificial intelligence has so far failed to reach the latter level. Ultimately, AI robots may, at most, possess “rational agency” in terms of functioning, based on both the standard conception of action (minimal conception) and the instrumental conception of “action.”

کلیدواژه ها:

Artificial Intelligence ، Ethical Agency ، Ethical Deadlock