Rule of Law in Algorithmic Government

Abstract

　The use of artificial intelligence (AI) in governments is examined in this article with a focus on the rule of law, transparency, and fairness. The subject matter revolves around algorithmic bias and hallucination in relation to automated decision-making processes. Machine learning bias makes it hard for AI to escape, so that human intervention is necessary to protect people’s rights. The difficulty of translating the statute’s language into the algorithm’s code is highlighted. Rules that differ from the original statutes may be prescribed by the algorithm due to the difficulty of translating. The rule of law is violated by it. The transparency of administrative processes that use AI is reduced by the difficulties of translation and the absence of AI literacy among officials and citizens, in addition to the black-box nature of machine learning. A quiet revolution has been initiated by the Digital Administrative Procedure Act of 2019 that encourages the sharing of personal information within the government. This paves the way for the era of electric government. A local statute has been passed in Kobe City that outlines how artificial intelligence is to be used in administrative processes.