Answers are grounded in this paper and cite their sources.
Generative artificial intelligence can describe a religiously associated object while making unsupported claims about the people who keep or use it. In inhabited heritage, this movement from visible form to religious attribution risks converting material resemblance into statements about contemporary practice, affiliation and recognition. This article develops a typed-claim audit distinguishing visible observation, bounded practice, identity attribution and religious recognition. Identity and recognition claims have restricted admissible forms: attributed reports of self-identification and recognition with explicit subjects, times and scope. The framework specifies source eligibility, records support and counterevidence separately, and distinguishes evidential support from delegation to speak on another’s behalf. Three analytical results follow under stated assumptions: identical visual records can underdetermine religious attribution; additional relevant counterevidence can withdraw admission; and an admission audit detects at least as many violations as an audit of exclusive evidential support on the same assertions. A fully disclosed eight-request constructed example demonstrates why assertion-level evidence-violation and admission-violation rates should be reported alongside answerable-claim coverage and response-level violation incidence. The latter remains unchanged when admissible assertions are appended to an already violating response. A research agenda specifies principles for future case documentation, controlled comparisons and independent coding. No field observations or model-performance measurements are reported. The contribution is an inspectable account of the evidential and representational limits of AI-mediated religious knowledge production, with source selection and reference judgements remaining open to contestation.
No takes yet. Share an insight, caveat, or question.
Lina Zhao (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: