Enhancing access to justice through chatbot technology: on the effectiveness of artificial intelligence in the delivery of legal advice

Many people are interested in how effectively robots and AI can enhance access to justice.This study examines Kazakhstan's digital court system and explores how large language models (LLMs) and question-answering (QA) tools can improve the quality, usability, and efficiency of legal services.The models -LLaMA-3B, GEMMA-9B, Phi-14B, Qwen-7B, and Mistral-7B -were trained and tested using data from leading Kazakh legal websites such as Adilet, Zqai, and Gov.kz, evaluated through ROUGE, METEOR, and expert review metrics.Among them, GPT-4o-mini achieved the highest accuracy and usability scores, demonstrating strong proficiency in the Kazakh language.The findings show that, in our study sample, AI-driven chatbots inspire more user trust than rule-based systems, indicating potential for boosting public confidence in digital judicial services.Overall, the study shows that LLM-powered legal advisory systems have the potential to advance sustainability, equity, and accessibility in national legal systems, which calls for further extensive research.