Editorial: AI-driven XR applications in cultural heritage: a new era of interactive museums

The proliferation of virtual and augmented reality technologies in cultural institutions has been described as heralding a paradigm shift, in which the ‘environment’ and ontology of cultural interaction are transformed, reshaping the dialectics of the ‘observer-artefact’ relationship (Shu, 2026). This formulation captures precisely what distinguishes the present moment from earlier phases of museum digitisation. Parry (2007) had already identified the sector’s initial encounter with computing as a set of technologies of change, while Cameron and Kenderdine (2007) argued that digital mediation reconfigures the ontological status of the cultural object itself. What is new is the extent to which this reconfiguration is now driven not by static digitisation but by extended reality (XR) and artificial intelligence (AI) systems that actively decide how, when, and to whom cultural content is delivered.

This shift demands sustained theoretical grounding, since technological capability continues to outpace curatorial and philosophical reflection. Virtual museology has traditionally rested on converting artefacts into networks of digital replicas (scanned, reconstructed, or born-digital) that circulate independently of their physical referents. What began as a hybrid, “phygital” design paradigm (Nofal et al., 2017; Nofal et al., 2020), in which digital layers supplemented tangible collections, has increasingly given way to fully virtual horizons in which the physical referent becomes marginal or absent altogether. This trajectory carries a corresponding shift in the unit of design: where curatorial practice once organised itself around the object, XR-mediated and AI-personalised experiences increasingly organise themselves around the embodied visitor. Whatever narratives, media, and channels this proliferation makes available, the cultural object’s significance must remain legible across them - a condition that is philosophical and ontological before it is technical.

The Metaverse extends this trajectory further, promising accessibility and inclusion unconstrained by geography or institutional gatekeeping. Yet full virtuality also produces a countervailing tendency: as the museum dematerialises, human interaction is not eliminated but relocated (De Marco, 2024), reasserting itself as a necessary complement to storytelling delivered through an otherwise non-human medium. This tension is sharpened by the uncertain role AI is being asked to play. AI-driven personalisation and affective computing systems increasingly claim the capacity to infer visitor states and adapt interpretive content in real time (Iakovaki et al., 2026), raising a question this Research Topic was convened to address: would it constitute a genuine improvement for AI to assume responsibility for curatorial intent, or would this represent a further instance of commercial and infrastructural interest encroaching upon a domain historically governed by scholarly judgement? The four contributions gathered within the Research Topic approach this question from complementary theoretical, infrastructural, and empirical perspectives.

Song and Evans provide the Research Topic’s theoretical foundation, applying Heidegger’s ontology of “the thing” and the fourfold to argue that artefacts acquire significance through gathered relations rather than physical properties alone, establishing curatorial intent as the condition under which XR-extended narratives remain legitimate.

Çakmak and Çekmiş operationalise this concern empirically, deploying an AI Director that adapts pacing and feedback to inferred visitor profiles while fixing artefact identity and narrative structure by design; offering the Research Topic’s clearest evidence that AI improves curatorial practice only when curatorial intent is engineered as a structural constraint.

Cunha et al. examine the opposite pole, documenting Museu XYZ’s blockchain-native, decentralised curatorial model, which substantiates the Metaverse’s accessibility promise while simultaneously illustrating the commercial entanglement flagged at the outset, given its dependence on cryptoart and NFT market logics.

Furusawa et al. demonstrate that aesthetic and emotional response in VR varies substantially with embodied locomotion and artwork scale, independently of visual attention, indicating that any AI-driven personalisation must ultimately model embodiment rather than preference alone.

Considered together, these contributions trace an arc from ontological theorisation, through constrained and unconstrained computational implementation, to embodied empirical evaluation, converging on the conclusion that the significance of cultural heritage in immersive environments remains contingent on embodied human experience rather than technological sophistication alone.

Themes explicitly invited by the call (i.e., gamification, XR-based museum games, and virtual humans or AI avatars) find plausible testbeds within the environments this Research Topic has described, though the Research Topic points to no single verdict. Decentralised, community-curated virtual museums, where curatorial authority is already distributed across artists and algorithmic collaborators, lend themselves to cooperative or competitive gamified mechanics, provided reward structures reinforce rather than dilute their non-hierarchical authorship. AI-directed installations anchored to physical heritage sites, whose real-time visitor profiling already differentiates pacing and feedback, sit close to museum-game mechanics in all but name; formalising this proximity would require explicit objectives and scoring without loosening the curatorial constraints that currently prevent drift into arbitrary content. Virtual humans and AI avatars could operate at either pole: as mediating presences restoring human-facing interaction to fully virtual spaces, or as embodied companions in immersive art-viewing environments, where their own embodiment and proximity would need the same careful design already shown to shape aesthetic response.

Beyond these specific themes, the broader opportunity this Research Topic points toward is a cultural heritage sector in which digital artefacts are treated not as static surrogates for physical originals but as first-class objects of curatorial, educational, and social interaction in their own right. Realising this opportunity will require future research that moves from single-institution deployments toward comparative studies across curatorial models (decentralised and centralised, physically anchored and fully virtual), incorporating non-adaptive baselines, longitudinal evaluation across repeated visits, and sustained critical attention to the economic and regulatory structures within which AI-driven curatorial systems are increasingly embedded. As these technologies mature, the philosophical, infrastructural, and embodied dimensions addressed in this Research Topic will remain essential to ensuring that innovation in cultural mediation serves curatorial and public interest rather than supplanting it.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited and reviewed by: Domna Banakou, Ecole Nationale Superieure d'Arts et Metiers - Campus d'Angers, France