Community-Grounded Authenticity at Bamiyan: Afghan Perspectives on AI-Mediated Heritage Reconstruction

Digital and AI-mediated reconstructions of destroyed heritage monuments have advanced in archaeological modelling, prompt engineering and expert-guided evaluation, yet the Bamiyan literature paid limited empirical attention on how Afghan respondents themselves define authenticity, acceptable reconstruction, and the representation of absence. This article addresses that gap through a questionnaire study of sixty-two Afghan respondents and situates the findings within the authors’ earlier reconstruction workflow. Support for reconstruction was high, but respondents did not equate acceptable reconstruction with visual replacement alone: alongside material fidelity, they prioritised historical context, cultural meaning, associated skills and knowledge, and selected sensory elements. Concerns about AI centred on the oversimplification of cultural meaning, while views on the inclusion of Hazara voices indicated continuing uncertainty about reconstruction governance. No meaningful differences emerged across the demographic sub-groups tested. The article argues that technically plausible and expert-validated reconstruction remains incomplete unless community-grounded authenticity is made visible in the reconstruction process. It accordingly proposes that reconstruction outputs published from AI-mediated workflows should record provenance, rupture, and governance visibility as minimum publication conditions. Preprint notice: This is the Author’s Original Manuscript/preprint version of a manuscript currently under consideration by International Journal of Heritage Studies. It has not yet been peer reviewed to completion, formally accepted, or published as the Version of Record. If accepted, this Zenodo record will be updated to link to the final published article.