Advancing AI-Driven Governance in Qatar’s Charitable Sector: A Trustworthy Arabic LLM for Regulatory Compliance

AI and NLP have transformed governance and compliance, yet Arabic-language applications remain underdeveloped, limiting legal interpretation and verification. Existing LLMs lack domain-specific training on Arabic legal and charitable documents, leading to inconsistencies and reduced trust in AI-driven decisions. This research develops an AI model tailored for regulatory compliance in Qatar’s charitable sector, integrating advanced Arabic language understanding with uncertainty detection and knowledge enrichment techniques. The proposed model improves accuracy, efficiency, and trust in AI-generated outputs for Arabic legal texts. It represents the first Arabic-focused LLM for regulatory compliance, incorporates mechanisms to mitigate AI hallucinations, and supports Qatar’s Vision 2030 by advancing digital governance. By enhancing legal compliance, AI governance, and transparency, this research positions Qatar at the forefront of Arabic AI innovation.

Hamad Bin Khalifa University (HBKU)