REGULATING AI-SUPPORTED LEGAL GUIDANCE IN DIGITAL LEGAL EDUCATION PLATFORMS: A RISK-BASED FRAMEWORK FOR YOUTH LEGAL LITERACY

Abstract

Artificial intelligence (AI) can expand access to legal knowledge by making legal information more searchable, conversational, and responsive to user needs. However, as digital legal education platforms begin to provide fact-specific explanations, personalized guidance, or recommendations, the boundary between legal education, legal information, and individualized legal advice becomes increasingly difficult to maintain. This challenge is particularly important for children and young users, who may be more vulnerable to legal error, over-reliance, opaque personalization, and unnecessary processing of personal data. This article develops a risk-based framework for AI-supported legal guidance in digital legal education platforms. Using doctrinal, comparative, and design-oriented methods, it examines the EU Artificial Intelligence Act, the Council of Europe Framework Convention on Artificial Intelligence, UNESCO and UNICEF guidance, the OECD AI Principles, professional ethics guidance, empirical research on legal hallucinations, and relevant Uzbek legislation and policy. The analysis also draws on the scientific-practical concept of the “Huquq maktabi 2.0” project. The article proposes four functional risk tiers: educational explanation, legal navigation, personalized legal guidance, and high-stakes legal advice. It argues that safeguards should increase in proportion to personalization, legal consequence, and user vulnerability. The proposed model combines authoritative-source grounding, transparency, human oversight, data minimization, age-appropriate design, referral mechanisms, and redress. Rather than relying on a binary distinction between “information” and “advice,” the article advocates a functional approach capable of supporting legal literacy while reducing the risks of unsafe automated legal guidance.