Generative AI in Shariah Advisory in Islamic Finance: An Experimental Study

Abstract

Islamic banks adhere strictly to Shariah principles in all operations and offerings, creating a demand for experts proficient in Shariah-compliant banking and finance. This study investigates the potential of using generative AI, specifically Large Language Models (LLMs), to provide Shariah advisory in complex Islamic financing scenarios. The research evaluates general purpose public LLMs—such as ChatGPT, Gemini, and Meta AI—on their ability to identify Shariah compliance issues, cite relevant Shariah references, and propose actionable guidance for resolving compliance challenges in complex Islamic financing scenarios. A qualitative approach was used. The experimental study involved creation of complex, realistic financing scenarios requiring Shariah opinion, expert consultation from real Shariah scholars on those issue, and then soliciting and evaluating Shariah opinions from LLMs. Results demonstrate their competence in identifying compliance issues and providing Shariah references, indicating a foundational understanding of Islamic finance. However, their inconsistent ability to offer actionable guidance highlights gaps in handling complex scenarios. While these tools can enhance efficiency in Shariah advisory by streamlining routine tasks, their limitations underscore the need for human oversight and domain-specific refinement. The findings pave the way for developing purpose built models and a frameworks to integrate AI into Islamic finance ethically and effectively.

Keywords

• generative ai
• large language model
• islamic banking
• shariah advisory.