Digital Sacred Texts, Algorithmic Interpretation and Religious Literacy: A Theoretical Framework for Scripture Engagement in the Age of Large Language Models

Abstract

Large language models (LLMs) now explain verses, summarise books of scripture, locate hadith and translate sacred texts on demand, and believers increasingly consult them in place of commentaries, teachers and communities of interpretation. Existing studies have assessed LLM outputs on particular texts and documented the general problem of fabricated content, but they have not offered a framework that distinguishes the different things readers do with scripture, nor connected these questions to religious literacy or to African interpretive traditions. This conceptual study addresses that gap. Its objectives were to define distinct modes of scripture engagement, to analyse the risks and possibilities LLMs bring to each, and to derive implications for religious literacy with attention to African hermeneutics. The study adopted a conceptual research design based on structured analysis of published literature. The population comprised peer-reviewed journal articles published between 2023 and 2026 on AI and sacred texts, the epistemology of LLMs, African biblical hermeneutics and scripture engagement, religious literacy, and reading and formation. Through purposive, criterion-based sampling, 30 articles were selected, each verified against its Crossref publisher record and screened for retractions. Data were analysed through inductive classification, mapping of documented LLM capacities onto each mode, and derivation of propositions. The analysis distinguishes three modes of scripture engagement: retrieval, interpretation and formation. LLMs can assist retrieval when grounded in authoritative sources, though fabricated citations remain a risk; they tend to flatten interpretation by reproducing dominant readings, lacking interpretive method and carrying bias; and they cannot participate in formation, which they may displace. For African traditions the risk is sharper, because LLMs are strongest in the languages and readings that mother-tongue and contextual hermeneutics have sought to decentre. The study recommends teaching mode awareness and verification as part of scriptural literacy, sustaining slow and communal reading, designing scripture tools that cite sources and show interpretive range, and developing AI resources in African languages with African interpretive communities.