Across Europe and globally, cultural institutions are rapidly integrating artificial intelligence, immersive media, and networked infrastructures into heritage and curatorial practice. These systems are typically evaluated through a functional paradigm that prioritises responsiveness, efficiency, and seamless optimisation. This paper argues that such criteria are insufficient. Digital mediation is not just reactive delivery but is a performative configuration of computational systems stage encounters through which cultural knowledge is structured, authorised, and experienced.

The paper introduces the Scene–Actor–Script–Audience (SASA) model as a transferable methodological framework for analysing and designing technologically mediated environments. Rather than focusing solely on data accuracy or system responsiveness, the model foregrounds how spatial settings, algorithmic logics, institutional authority, and embodied participation interact to produce meaning. Central to this approach is a distinction between reactivity and performativity: systems may function reliably without continuously responding on demand, instead shaping temporal rhythm, attention, and interpretive possibility.

The framework is applied to two completed case studies in Chinese cultural contexts: a projection intervention within a UNESCO World Heritage garden in Suzhou and an augmented reality installation in a transcultural architectural environment. These projects are positioned not as peripheral examples but as methodological sites of learning. By working within non-Western spatial philosophies that privilege relational pacing and environmental primacy, the cases challenge the global homogenisation of audience-facing technologies in cultural institutions.

As generative AI increasingly contributes institutional voice across multilingual and transnational infrastructures, questions of authority intensify. Linking digital environments across Europe requires more than technical interoperability; it demands attention to how mediation performs across borders. By modelling digital systems as staged actors rather than neutral tools, this paper offers digital humanities a framework for designing and governing AI-mediated cultural infrastructures in ways that remain situated, reflexive, and accountable.