Artificial Intelligence for Sustainable Cultural Heritage: Practical Guidelines and Case-Based Evidence

Abstract

The sustainable preservation of cultural heritage, as articulated in Sustainable Development Goal (SDG) 11.4, requires strategies that not only safeguard tangible and intangible assets but also enhance their long-term cultural, social, and economic value. Artificial intelligence (AI) and digital technologies are increasingly applied in heritage conservation. However, most research emphasizes technical applications, such as improving data accuracy and increasing efficiency, while neglecting their integration into a broader framework of cultural sustainability and heritage tourism. This study addresses this gap by developing a set of practical guidelines for the sustainable use of AI in cultural heritage preservation. The guidelines highlight six dimensions: inclusive data governance, data authenticity protection, leveraging AI as a complementary tool, balancing innovation with cultural values, ensuring copyright and ethical compliance, long-term technical maintenance, and collaborative governance. To illustrate the feasibility of these guidelines, the paper analyses three representative case studies: AI-driven 3D reconstruction of the Old Summer Palace, educational dissemination via Google Arts & Culture, and intelligent restoration at E-Dunhuang. By situating AI-driven practices within the framework of cultural sustainability, this study makes both theoretical and practical contributions to heritage governance, to enhance cultural sustainability commitments and align digital innovation with the enduring preservation of humanity’s shared heritage, providing actionable insights for policymakers, institutions, and the tourism industry in designing resilient and culturally respectful heritage strategies.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 29 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.