BIASED ALGORITHM, ARTIFICIAL INTELLIGENCE AND ITS CRIMINAL LIABILITY FROM PERSPECTIVE OF RETRIBUTIVISM

At the moment, robots are not legal persons and therefore cannot be held liable for the criminal offenses they committed. In this article, I would like to discuss the legal status of Artificial Intelligence (“A.I.”) and further examine the issue of punishing them when they commit crimes against humans. Starting with the Part I Introduction, the Part II of the article provides a brief discussion of the relationship between A.I. algorithm and human’s “greater happiness” from utilitarian perspective. In Part III, the types of big data discrimination and its criminal liability resulting from such algorithm will be examined, specifically on police robots. Part IV offers an in-depth analysis on A.I.’s legal status- who will become responsible for the crimes committed by A.I.? Some scholars in the field have endeavored to make comparison analysis in various ways: A.I. and corporate legal entity, A.I. and agency liability, or A.I. and dangerous animals, etc. I aim to provide somewhat different point of view from so-called “retribution gap” and the theory will be further delineated in Part V. Lastly, the conclusion suggests that new dominant legal paradigm defining the scope of A.I. rights and obligations is indeed necessary and that it will eventually replace the existing one.