Can LLMs Write Faithfully? An Agent-Based Evaluation of LLM-generated Islamic Content

Abdullah Mushtaq, Rafay Naeem, Ezieddin Elmahjub +5

Large language models are increasingly used for Islamic guidance, but risk misquoting texts, misapplying jurisprudence, or producing culturally inconsistent responses. We pilot an…