Artificial Intelligence in Sacred Text Analysis: A Survey of Methodologies and Applications

Abstract

The past decade has witnessed a remarkable convergence of artificial intelligence (AI) and the scholarly study of sacred texts. This paper offers a critical survey of AI-based sacred text analysis, drawing on studies involving the Bhagavad Gita, the Quran, and the Bible. It reviews approaches ranging from lexicon-based sentiment analysis and topic modeling to transformer-based models such as BERT, and it examines where these methods succeed and where they fall short. A central argument of the paper is that sacred texts cannot be treated as ordinary sentiment datasets because they are culturally embedded, symbolically rich, and theologically nuanced. Rather than claiming empirical validation beyond what is reported in the literature reviewed, this paper synthesizes prior work and proposes a conceptual framework built on interpretive partnership, methodological transparency, and domain-informed training. The study contributes a more careful and responsible perspective on the use of AI in spiritually sensitive domains.