Legal Measures for Regulating the Use of Artificial Intelligence in the Thai Justice System

Abstract

This article examines the role and applications of Artificial Intelligence (AI) in the justice system, including the use of Machine Learning and Natural Language Processing (NLP) systems to assist with legal research, contract review, and case outcome prediction. However, the use of such technologies raises significant legal issues, particularly when AI malfunctions or generates false information (AI Hallucinations), undermining trust and causing harm to the parties involved. The study finds that Thailand currently lacks specific legislation to regulate the use of AI. Relying on the comparative application of the Product Liability Act B.E. 2551 (2008) creates interpretive problems regarding whether AI is considered a "product" and raises difficulties concerning the burden of proof for third-party victims. A comparison with foreign jurisdictions reveals interesting legal measures, particularly the European Union's AI law (EU AI Act), which utilizes a "risk-based approach" to classify AI into four levels with proportional regulatory measures. Furthermore, there are state-level laws in the United States, such as Colorado, which focuses on regulating high-risk AI to prevent discrimination; Utah, which mandates that professionals disclose the use of AI to consumers; and California, which has laws that protect transparency and prevent threats from Deepfakes. Therefore, this article recommends that Thailand urgently enact a dedicated Act to regulate AI using a risk-based classification model. This should be coupled with the establishment of a regulatory body empowered to impose legal penalties for enforceable regulations. This will serve as a crucial mechanism for protecting citizens' rights and elevating the Thai justice system to international standards.

 Keywords: Artificial Intelligence, Justice System, AI Regulation, Product Liability Act