Abstract
This research looks at how different areas of law handle the application of AI to the adjudication process and how it affects the protection of the right to a fair trial. A independent, competent, and impartial court must be able to hear a case in a timely manner in a public setting without unnecessary delay; however, the author argues that this is not always achievable with AI-powered information systems. Considering the unique features and goals of criminal, civil, and administrative court procedures, the article addresses these concerns in connection to each. As part of his analysis of the research challenge, the author mentions the right to a cursory review of the case and its potential impact from AIbased information systems. The research makes use of a critical literature review approach in addition to a methodology for studying theories of law and the law as it is in effect. The favoured findings were obtained by employing various classifiers, including the AE, MLP, RNN, VAE, and AE-MLP. While AE achieved 90.38% and MLP 94.17%, AE-MLP achieved 97.62%. When compared to other models, AE-MLP attained a higher level of accuracy, while VAE achieved 88.76% and RNN achieved 96.43%. They further show that, when predicting case outcomes using only the semantic bias categorisation in judicial judgements, we achieved a relatively high classifying score.
Keywords: Artificial Intelligence (AI), Auto-Encoder (AE), Judicial Proceeding, Multi-Layer Perceptron (MLP), Principle Component Analysis (PCA).