1. Introduction: a nation and its twin
In 2021, Tuvalu’s Minister for Justice, Communication and Foreign Affairs, Simon Kofe, made headlines when he addressed the UN climate change conference. “We are sinking,” he told the world, “but so is everyone else” (Kofe Reference Kofe2021). The filmed speech received a punchy twist when the camera zoomed out, revealing Kofe standing knee-deep in his suit and tie, where he was delivering his speech from a lectern set up in the sea. “We will not stand idly by as the water rises around us,”Footnote 1 he promised, and he certainly did not do so going forward. The following year, 2022, the minister of a country that described itself as a place almost no one knows sent another video to the COP27 climate change conference. This new video was highly produced, yet something about it felt odd. It was almost as though it looked too realistic. In it, Kofe appears again, composited into a hyperreal, digitally simulated rendering of Teafualiku Islet, the smallest of Tuvalu’s nine islands and the first expected to disappear due to sea level rise. From Teafualiku’s digital replica, Kofe announced the nation’s “Future Now” plan to become the world’s “First Digital Nation,”Footnote 2 using a digital twin as a virtual replica of a physical system that was continuously updated with data from its counterpart. Often described as the canary in the coalmine of climate change, the former British colony will become uninhabitable for its 11,000 people long before its complete sinking. While the politically charged term “First Nation”Footnote 3 very intentionally does not have a fixed legal definition, the ambition to become the first digital nation on the metaverse by using a digital twin of the island has produced a number of technically ambiguous theoretical inquiries with regard to cultural preservation in the age of deep learning and generative artificial intelligence (AI).
This article examines Tuvalu’s digital twin – designed for preservation and archiving of a nation – not simply as a technical solution, but as an emerging governing paradigm that is grounded in pattern discovery, unstructured data and domain agnosticism (Amoore Reference Amoore2025; Borbach et al. Reference Borbach, Chun and Thielmann2025). Navigating this intersection of simulation and statecraft, the article asks what is being preserved by using this process, by whom and for whom? It will explore how Tuvalu’s state-scale digital twin extends archival governance into predictive and interventionist computational regimes. It will furthermore offer the term “knowledge islandization” to capture a colonial epistemic technique that extracts and fragments relational worlds into discrete and governable units under the guise of preservation. This has an older history but is clearly intensified in the Digital Twin. Throughout, I will use performance theory to read the modern state archive and the emerging digital twin and analyse both as highly performative agents. These agents actively refine the perceptions of citizenship, local knowledge and statehood. I conclude by suggesting reimagine preservation practices that do not replicate the reductive logics of colonial islandization.
1.1 Article structure and methodology
The article opens with a technical and critical introduction of the digital twin, an overview of Tuvalu’s path to digital nationhood, and the introduction of the Tuvalu twin through the framework of an “Is/As Archive.” I keep the two terms, “digital twin” and “archive,” in a deliberately close, but not interchangeable, relation, and I write across three distinct and sometimes overlapping registers. Firstly, I analyse instances when the twin is clearly an archive, using archival theory. Secondly, I describe how the twin functions archivally, when specific parts of the project perform archival work through the accumulation, certification and governance of records. Finally, I highlight how and when the digital twin exceeds classical archival logics and becomes a device for intervention through operations such synchronized modelling, simulation and prediction. This distinction avoids collapsing the twinning into storage or data collection functions, while keeping archival critique as a tool through which to read the digital twin.
In what follows, taking inspiration from the atoll state, the analysis is organized into four “islands” – Rescue, Archive, Proxy and Data – each of which addresses the research question by examining a different facet of the wider transformation by which the island’s culture, environment and sovereignty are reconfigured in the preservation process. Rescue Island reflects on the notion of rescue as a colonial practice and its implantation in the technosphere and demonstrates how the language of rescue is itself a cultural and moral asset for external actors. Archive Island draws on archival studies to analyse the twin through archival logics, tracing how accumulation becomes epistemic power, and where twinning begins to exceed archival custody. Proxy Island examines how the digital ex situ nation depends on a dense ecology of human and nonhuman proxies for the maintenance, governance, authentication and circulation of its different knowledges. Finally, Data Island explains how the twin’s capacity to simulate realities and monitor systems in real time transforms the island into a manageable data asset controlled by multiple autonomous systems, thereby organizing value, authority and jurisdiction in ways that risk renewing colonial structures.
Examining the technical accounts and cultural critiques of the digital twin, alongside the concepts of archival theory, through the lens of performance theory allows me to analyse the social and political agencies involved in the twining process. This approach is extended through offering empirical evidence of Tuvalu’s digital transition, still in its early stages, and through observations of media products, including websites, demos and clips, that present Tuvalu’s digital transformation process from 2022 to 2025. These observations make it possible to analyse the contingent promises and claims made for Tuvalu’s digital twin platform, and engage with its articulation of the mission as well as the hidden assumptions that are clearly inscribed in what the twin is designed to achieve. In addition to these observations, conversations with individuals who are directly involved in the development of the cultural repository provide a more situated sense of the process of building what is presented therein as the first digital nation.
I have never been to Tuvalu, and I assume many of my readers have not either. My understanding comes from a distance, shaped by ongoing theoretical and artistic research on the digital afterlives of cultural objects threatened by mass destruction. I approach Tuvalu not as a fixed case study but as a constellation of experiences, temporalities and settings that demand new ways of rethinking archival practices under conditions of planetary precarity.
I write as a practice-based artist-researcher working in and with archives, performance and media. This text forms part of a wider artistic and theoretical research project on Tuvalu, spanning speculative writing, lecture performance and sound-video installation. I treat institutional documents, interfaces, announcements and protocol texts as performative artefacts that do not merely describe “digital transition” but actively compose its authority, publics and futures. This is not an ethnography of Tuvaluan lifeworlds. The community roundtable protocols I cite are translated records shared with me by organizers and participants,Footnote 4 read here as textual traces of governance, disagreement and consent-making. From this position, I bring critical media analysis into dialogue with archival theory and performance studies to ask what kinds of memory, evidence and sovereignty are being produced, and at what cost, when preservation is refigured as an operational, data-driven infrastructure. In doing so, I extend Annet Dekker’s call to rethink archival practices through ecological imperatives, fluidity, ephemerality and collaboration, while situating this rethinking within wider efforts to challenge the colonial logics that have long shaped heritage preservation.Footnote 5
2. Digital twin: technical and critical introduction
A digital twin is often described as the virtual replica of a physical product, system or environment that is continuously updated with data from its counterpart (Wang et al. Reference Wang, Yang and Zhu2021). The concept was first formalized by Michael Grieves in 2002 as a “virtual instance of physical assets capable of continuously mirroring them” (Grieves Reference Grieves2022), though the genealogy of digital twinning is often traced avant la lettre to NASA’s use of real-time spacecraft simulations during the Apollo 13 crisis (Glaessgen and Stargel Reference Glaessgen and Stargel2012). In recent years, digital twins have proliferated across industrial, engineering and environmental domains and have been promoted as tools for optimization, prediction and process control (Jiang et al. Reference Jiang, Yin and Li2021; Rathore et al. Reference Rathore, Shah and Shukla2021). Their rise is situated within a longer history of scientific representation tied to probability, visualization and statistical knowledge (Galison Reference Galison1997; Hacking Reference Hacking1990). In computer science discourse, these are described as “live copies” of physical systems, valued for accuracy, synchronicity and their capacity to integrate multimodal data streams, from sensors and LiDAR scans to text, images, audio and video (Amoore Reference Amoore2025; Mashaly Reference Mashaly2021). The novelty of a digital twin is in its ability to combine structured and unstructured data sources from the physical environment into a single model, thereby creating an apparently seamless copy that enables continuous transmission between the physical and virtual domains (Fuller et al. Reference Fuller, Fan and Day2020; Wu et al. Reference Wu, Zhang and Zhang2021). As Leighton Evans (Reference Evans2025) noted, the reflexivity and accuracy attributed to digital twins have made them increasingly attractive to cultural heritage projects. Over the past two decades, these twins have been increasingly used for the preservation of heritage sites, extending earlier experiments with virtual environments and digital reconstructions (Hutson et al. Reference Hutson, Weber and Russo2023). Tuvalu’s digital twin is expected to integrate satellite, drone and street-level imagery with Geographic Information System (GIS) to construct a three-dimensional (3D) metaverse environment supported by cloud-hosted governance systems and blockchain identification, while still incorporating real-time environmental data, AI-driven simulations and a hazard monitoring dashboard for clear prediction and decision-making. The national twin will be experienced via a virtual reality platform that will help displaced and future generations of Tuvaluans to “engage with each other in the Metaverse, in a virtual space that mirrors real life and preserves the shared language and customs” (Evans Reference Evans2025, p. 4500).
Next to technical literature, recent critical technology scholarship positions digital twins not as neutral simulations of physical processes, but as political technologies used for governance. Louise Amoore describes them as governing conduct “through accumulation and archiving,” shaping both “spaces of possibility and decision” while simultaneously foreclosing alternatives by dictating what the projected, possible futures are and which ones are implausible (Amoore Reference Amoore2025, p. 4562). Scholars emphasize that, although often framed as digital copies of the physical world, digital twins actually diverge sharply from being mimetic replications (Borbach et al. Reference Borbach, Chun and Thielmann2025; Korenhof et al. Reference Korenhof, Blok and Kloppenburg2021). Predictive systems can privilege correlation over fidelity, producing models that are “good enough” to act on rather than accurate representations (Chun Reference Chun2021), thereby advancing a mode of twinning that classifies, extracts and divides. For Amoore, twinning is an apparatus of intervention, prediction and action that marks a paradigmatic break from earlier models of simulation. In the age of AI, these digital twins thus function less as a mirror of the world than as a socio-political technology that archives, classifies and performs the world. They exemplify a technocratic paradigm (Borbach et al. Reference Borbach, Chun and Thielmann2025) that is premised on the imperative to render the world calculable by translating material processes into data environments in real time for prediction and intervention. They thus operate as entangled practices of decision-making, steering and governing (Dawkins and Kitchin Reference Dawkins and Kitchin2025; Korenhof et al. Reference Korenhof, Blok and Kloppenburg2021), while simultaneously forming spaces and building infrastructural environments (Horn and Richardson, Reference Horn and Richardson2025). Sustained by continuous data streaming between physical objects and their virtual counterparts, digital twins can collapse the distinction between representation and operation and produce futures as much as they claim merely to anticipate them.
2.1 Twinning Tuvalu
Tuvalu’s unique situation and possible digital future have drawn sustained attention across climate research and diplomacy (Mahony and Hulme Reference Mahony and Hulme2018; Mortreux and Barnett Reference Mortreux and Barnett2009; Saddington Reference Saddington2023), international affairs (Barnett et al. Reference Barnett, Farbotko, Kitara and Aselu2025; Dun et al. Reference Dun, Klocker, Farbotko and McMichael2023; Rothe et al. Reference Rothe, Boas, Farbotko and Kitara2024), law (Iyer and Schewel Reference Iyer and Schewel2024), Pacific studies (Kitara Reference Kitara2024; Lusama Reference Lusama2022; Suliman et al. Reference Suliman, Farbotko, Ransan-Cooper, McNamara, Thornton, McMichael and Kitara2019) and media and technology (Echavarria Reference Echavarria2025; Evans Reference Evans2025; Otero Verzier and Miller Reference Otero Verzier and Miller2025). At the crossroad of a disappearing locale and continuing sovereignty, the twin project combines two archival objectives: (1) preservation, a utilization known from other cultural heritage projects, and (2) actionable governance. It can thus be read as the first digital twin of an ex situ nation. The concept of ex situ nationhood, proposed by legal scholar Maxine Burkett (Reference Burkett2011), addresses the legal and political challenges of nationhood under climate-driven displacement, envisioning a continuity of sovereignty, identity and international recognition when land becomes uninhabitable; this status allows a dispersed population to deterritorialize a nation and continue exercising political and legal autonomy. In Tuvalu, the Falepili Union treaty, signed in 2023, allows for the migration of 280 people from Tuvalu to Australia each year, while also recognizing Tuvalu’s statehood will continue even if its land is inundated by climate-related sea level rises.Footnote 6 Building on Burkett’s general formulation of ex situ nationhood as a continuity of sovereignty without any inhabitable territory (Burkett Reference Burkett2011), later publications have looked to Tuvalu’s “First Digital Nation” as a testing ground for ex situ sovereignty (Rothe et al. Reference Rothe, Boas, Farbotko and Kitara2024; UNDP 2025).
It is very important to acknowledge that the twin is not fully realized yet, and the preparations for it are rather opaque. However, waiting for the twin is also performative. Following his two UN videos, Kofe appeared on the national website asking, “If you were about to lose everything, what is the one thing you would save?” (Kofe Reference Kofe2023). He then promises that Tuvaluans’ answers – artefacts, children’s language, elders’ stories and festival dances – will be digitized into a “digital ark,” a national archive built for crisis that invokes the Biblical flood and frames a mission “not just to rescue Tuvalu, but to save the world” (Kofe Reference Kofe2023). An accompanying animation visualizes the archiving process, from recipes and customs to songs, tools and wearable ephemera, making the custody goal explicit. In parallel, the national campaign made by Kofe (Reference Kofe2023) announces that “our digital migration has begun” as a message produced by the Australian ad agency, The Monkeys, widely circulated to signal momentum, even as the path to a digital state is only just beginning.
This narrative is now coupled with institutional infrastructures. About three months after Kofe’s virtual address, Tuvalu announced a partnership with the German Archaeological Institute (DAI) and the United Nations’ Rising Nations Initiative (RNI), a programme focused on Pacific atoll states that are facing catastrophic risk. The partnership is assigned to build the Tuvalu Digital Repository for Cultural Heritage, which was publicly launched in June of 2025. However, digital twin Tuvalu in its full realization is not simply a cultural repository, but rather a project based on different scale, ambition and ideology. If completed, all things under the Pacific sun are planned to be transformed into a data format and thus become quantifiable entities – physical territory, governing sovereignty and a record-keeping entity.
3. Is/as an archive
In his seminal work on performance studies, Richard Schechner introduced the concept of “is/as” – a framework for analysing performance as both an inherent and an applied phenomenon. For Schechner (Reference Schechner and Brady2013), anything can be examined “as” performance, meaning that performance theory can be used to critically engage with any subject matter. In contrast, something “is” performance when historical social context, convention, usage and tradition define it as such. Employing the “as/is” framework into archival context helps to better understand the relation between the digital twin and the archive as it shapes the production of knowledge, authority and statecraft.
Focusing on the infrastructural and spatial systems of statecraft (Easterling Reference Easterling2014), it is notable that the technological platform hosting Tuvalu’s digital state is designed to govern “through accumulation and archiving” and to shape “spaces of possibility and decision” (Amoore Reference Amoore2025, p. 4562). As Eun Seo Jo and Timnit Gebru (Reference Jo and Gebru2020) show, the archive is the earliest form of large-scale, collective human record-keeping, and the earliest archives have been state-instituted, with the purpose of governing the public through collecting individual, family, organizational, financial and governmental records. The digital twin from this perspective is a new form of national archive: a large-scale, machine-learning-driven, state-instituted repository of legal, cultural and environmental collective record-keeping. The digital twin governs through accumulation and privileges correlation over fidelity (Borbach et al. Reference Borbach, Chun and Thielmann2025; Korenhof et al. Reference Korenhof, Blok and Kloppenburg2021). Therefore, being archived in this way carries new and unforeseeable risks and invites a reconsideration of authoritative power and asks what that power reveals about the archive as a national apparatus of memory.
Crucially, the equation of the twin with the archive offers only a partial account of its function. Parallel to the register that shows the twin is an archive, I also read it as an archive and propose that the archive and the digital twin are social and political technologies. To do so, I build on existing critical literature that applied archival theory to engage with algorithmic decision-making (boyd and Crawford Reference boyd and Crawford2012), big data (Thylstrup et al. Reference Thylstrup, Agostinho, Ring, D’Ignazio and Veel2021) and machine learning datasets (Jo and Gebru Reference Jo and Gebru2020).Footnote 7 Building on debates associated with the archival turn in the humanities, recent scholarship argues that digital infrastructures reproduce enduring political and epistemological tensions concerning who may be seen, how materials are rendered legible and under what conditions knowledge circulates (Thylstrup Reference Thylstrup2022). From this perspective, digital archives do not constitute a qualitative rupture with earlier archival regimes but represent a quantitative transformation in scale, materiality and governance that demands renewed critical scrutiny (Agostinho Reference Agostinho2019; Agostinho, Dirckinck-Holmfeld, and Grova Søilen Reference Agostinho, Dirckinck-Holmfeld and Grova Søilen2019; Sutherland Reference Sutherland2017). Reading the Tuvalu twin as an archonic apparatus thus allows me to follow the “as/is” distinction without reducing the digital twin into the archive. It also exposes instances in which it exceeds archival logics, through the transformation of record-making into predictive intervention, and of data produced through datafication into a political and economic asset.
Why introduce performance theory to this case? Critical literature on the performative aspect of the archive (Borggreen Reference Borggreen2013; Schneider Reference Schneider2011; Taylor Reference Taylor2003) shows how collecting, sorting, recording, repeating, replacing and presenting, as well as concealing, erasing or omitting, are all part of the archival repertoire and crucial to its political existence. Parallel observations have been expressed by Borbach, Chun and Thielmann, who described digital twins as performative agents embedded in feedback loops between humans, machines, environments and algorithms (Reference Borbach, Chun and Thielmann2025, p. 4374). The performativity of the digital twin is still not easy to capture. Like the archive, it operates away from the public eye with limited access for a general audience. Both offer a kind of performance without spectacle, intensifying governance while reducing public visibility. Read through performance theory, the twin allocates attention and resources, normalizes classifications as operational thresholds and binds archival description to computational action (Borbach et al. Reference Borbach, Chun and Thielmann2025). In this sense, the archive and the twin function in similar ways in the production of knowledge, authority and statecraft. This is/as archive framework provides the basis for the Tuvalu-specific analysis that follows.
4. Knowledge islands
The four sub-sections below highlight the different aspects of the national twin is/as an archive, and ask how archival authority is reorganized, and how twinning pressures reshape what counts as record, proxy and asset. This structure underscores what I call the islandization of knowledge, a mode of archiving that extracts, fragments and repurposes knowledge under the guise of preservation. By that, I refer not only to a spatial metaphor but to a colonial epistemic technique that renders relational worlds into discrete and extractable units: islandization names a process through which territories, communities and knowledge systems are separated, miniaturized and made governable through extraction and partitioning. The origins of this logic lie in colonial practices that broke Indigenous worlds into administrable fragments. These included cadastral and shoreline surveys and ethnographic procedures that made Indigenous life legible for administration (Scott Reference Scott1998), as well as the archival practices such as registration of cultures, linguistic types, specimen objects and manageable micro-territories driven by the impulse to reorganize, capture, claim and convert a place into scientific, economic or cultural assets (Arvin Reference Arvin2019). Island studies scholars show how islandness has been repeatedly constructed through tropes of smallness, boundedness and otherness (Foley et al. Reference Foley2023), a construction that shaped the position of islands in relation to metropolitan states and empires and aligned with sociological inquiries into power and hegemony (Chandler and Pugh Reference Chandler and Pugh2021). Earlier accounts describe islands as concepts that dictate both unity and separation, inclusion and exclusion (Baglole and Weale Reference Baglole and Weale1973, pp. 105–06), and these accounts contributed to forms of island identity built on distinctions between islanders and those marked as others (Ronström Reference Ronström2021; Royle and Brinklow Reference Royle, Brinklow and Baldacchino2018). As a result, policies, management frameworks and governance arrangements for islands have often been shaped by outsider representations and claims to superior agency over local social or environmental futurity (Hau’ofa Reference Hau‘ofa2008).
The idea of knowledge islandization extends beyond island life. It operates as an enduring epistemic technology, manifested in the logic of the colonial archive and the museum collection,Footnote 8 and rearticulated through contemporary regimes of digital preservation. Pacific archivists have shown that digitization infrastructures often reproduce the same logics of colonial isolation, transforming relational cultural knowledge into discrete digital units organized through external metadata schemas, platform taxonomies and access protocols (Francis and Liew Reference Francis and Liew2009; Singh and Blake Reference Singh and Blake2012). Islandization of knowledge, therefore, names a persistent colonial sociotechnology (Feyrer and Sacerdote Reference Feyrer and Sacerdote2009; Jones Reference Jones2019) that moves between the geographical and the imaginary. It carves ocean worlds into separate units of analysis and governance and reproduces reductive and divisive colonial framings. When converted into digital infrastructures, separation becomes a governing principle for what can be known, stored and acted upon. Tongan anthropologist Epeli Hauʻofa had characterized these framings as a colonial habit that renders the Pacific as “islands in a far sea” (1994), defined by remoteness and discreteness.
Against this tradition of colonial knowledge islandization, island thinkers from Hauʻofa (Reference Hau’ofa1998) to Édouard Glissant (Reference Glissant1997) develop epistemic models that foreground circular and relational movement between land and sea and between islands. Gilles Deleuze (Reference Deleuze2004) and Jacques Derrida (Reference Derrida2011) have also employed the island as a figure that explicitly disrupts reductive frameworks of modernist and colonial mainland thinking. Thinking with islands and islandness offers alternatives to modernity’s linear narratives of progress and development, and coloniality’s models which administrate the world in extractable units, which are now widely recognized as drivers of climate catastrophes (Pugh and Chandler Reference Pugh and Chandler2021).
Knowledge islandization, understood as an enduring epistemic technology, persists across aid, environmental research and funding systems. Methods familiar from colonial collecting and archival practice that involve knowledge islandization now reappear in Tuvalu’s digital transformation, specifically in the digital twin. In what follows, I will show how these methods are enacted in dataset design and capture, in systems of classification and indexation, and in regimes of storage and access. In each case, processes that once reduced cultural and ecological worlds into extractable fragments re-emerge as computational protocols that organize what can be known, preserved and made actionable within digital infrastructures.
4.1 Rescue island
The vocabulary of crisis that frames Tuvalu’s environmental precarity is mirrored in the technological solutions proposed for its digital future. This section thus argues that the grammar of risk and rescue does not merely describe crisis but also authorizes selection, extraction, classification and custody. Read using this grammar, the digital twin is legible as an archive that is assembled under emergency, and it converts the island into a cultural and moral asset for those who do rescue it.
The weather makes and unmakes Tuvalu. It is the least visited country in the world, where the local airport is also used as a football field that is also routinely subjected to oceanic “king waves” that turn its single landing strip into a water slide. The weather displaces land, damages plantations and separates families. It is physical, social, spiritual, political and emotional – every day its density increases, and it is becoming ever more pervasive. It is this context and milieu with which Tuvalu’s citizens must continually contend. To reflect on Christina Sharpe’s (Reference Sharpe2016, p. 134) writing about the material and political agency of the weather, Tuvaluan people are constituted through and by their continued vulnerability to this overwhelming force. As a small island nation, Tuvaluan culture is crucially centred around oceanic weather, but in recent decades that same weather became “the problem.”
In climate research, the shoreline is described as a “riskscape” (Yarina and Takemoto Reference Yarina and Takemoto2017), a term that previews and then performs the threat of danger. For locals, however, the shoreline is a porous, dynamic zone, interconnected with other places (Arnall and Kothari Reference Arnall and Kothari2020). This concept of an “edge” lends itself to performative technocratic solutions like seawalls, reinforcing an illusion of control over dynamic atolls (Saddington Reference Saddington2023; Yarina and Takemoto Reference Yarina and Takemoto2017). Tuvalu’s weather then becomes a stage for the geopolitics of fear, enacted through alarmist geographies of catastrophe (Chaturvedi and Doyle Reference Chaturvedi and Doyle2010). This shift from relation to segmentation is not incidental; it is central to what I describe as “knowledge islandization,” as it administers the shoreline, a key site of livelihood and circulation, through partitioning and separation. The performativity of language represents an interplay between politics, aid, technocratic solutions, media representations and global aspirations, and converts situated lifeways into standardized fragments that lend themselves to both classification and custody. In a series of informal interviews with researcher Giulia Di Fonzo, Tuvaluans participants emphasized how much Western research literature discussing sea level rise, climate mobility and relocation centres on victim narratives, thus silencing and generalizing local voices (Di Fonzo Reference Di Fonzo2025, p. 29). Di Fonzo mentions that participants in her study identified the neocolonial orientations that turn Tuvalu into a climate-change laboratory, utilizing a “canaries in the coal mine rhetoric” (Gemenne Reference Gemenne2010) that “creates a hopeless and victim-oriented narrative that all participants expressed an urgency to push against in their work” (Di Fonzo Reference Di Fonzo2025, p. 29). Other assessments ranged from describing the initiative as a “flash in the pan” to seeing it as a “valuable asset” for managing climate impacts (Di Fonzo Reference Di Fonzo2025, p. 29), reflecting the way the project is folded into the wider structure of climate finance.
Moving from physical geographies to the digital, crisis grammar endures. Examples from digital heritage show how victim-oriented narratives, paired with the rhetoric of “rescue,” often travel from climate discourse into technological preservation through a vocabulary of assistance and provision that is seductive but also misleading. In the work on the digital preservation of sculptures destroyed by the Islamic State of Iraq and Syria (ISIS), the artist, Morehshin Allahyari, observed how non-governmental organizations (NGOs), governments and private firms rely on a language of “assisting” or “providing technology,” thereby promoting a narrative of “collective reconstruction” that presumes a paternal “we” that is safeguarding “their” culture (Allahyari Reference Allahyari2019). This critique resonates with heritage scholarship showing that preservation privileges historically sedimented grammars of value, including professional expertise, material authenticity and state authority, while often subordinating alternative epistemologies and community claims (Smith Reference Smith2006; Thwaites et al. Reference Thwaites, Turner and Ireland2025). In Tuvalu, Kofe’s videos address “the world,” saying that Tuvaluans “can’t do it alone” and exemplifying how the language of allies, such as “collective reconstruction,” is deployed and even internalized. Yet, the videos leave open the question of what exactly cannot be done alone, maintaining ambiguity in the call for global support. Rothe et al. (Reference Rothe, Boas, Farbotko and Kitara2024) suggested that the video addresses two audiences at once: Internally, it sustains Tuvaluan identity through virtual relations; externally, it legitimizes a deterritorialized Tuvalu on a global stage and exemplifies how the discourse of “rescue” has been internalized and extended from climate to culture. For example, the image of the ark binds the state repository and survival mechanism. Such discourse sustains the division between “we” and “ours” and “you” and “yours,” disregarding the agency of those whose histories are being archived and producing what has been described as forms of “violent care of conservation” (van Dooren Reference van Dooren2015). Taken together, the rhetoric of “rescue,” “crisis” and “care” risks creating a new form of digital violent care that, under the guise of aid, replicates coloniality of power.
Material arrangements also make the violent aspects of digital preservation visible. As Marina Otero Verzier and Dan Miller noted, Tuvalu “exists in digital traces, in fragments anywhere the Internet would transmit them” (Otero Verzier and Miller Reference Otero Verzier and Miller2025, p. 189), with resolution completely dependent on storage materiality: “Tuvalu may be in Ashburn Virginia. It may be in the Amazon Data Realty Storage Center that hosts Tuvalu’s website and part of their state dot tv domain” (Otero Verzier and Miller Reference Otero Verzier and Miller2025, p. 189). If storage and computation both rely on external cloud infrastructures, then selection, description, access control and even uptime depend on institutions beyond Tuvalu, meaning that parts of the national record are preserved and rendered actionable elsewhere, under the protocols of service providers.
What is rescued, therefore, is not Tuvaluan life worlds, but a portable origin story that fits the already designed requirements of standardization and custody, reproducing islands as extractable units that are then distributed across data centres and cloud storage solutions. Here, preservation becomes a material and geographical operation through which territories and situated knowledge systems are rendered governable by extraction, partitioning and redistribution.
Read through the concepts of risk, rescue and fix, this mode of preservation begins to serve a specific governmentality, one that becomes especially clear in ex situ scenarios, as I will show later, where natural and cultural knowledge is broken into “building blocks” for rebuilding or for addressing unknown future problems (Penrose, Harrison and Breithoff Reference Harrison, DeSilvey, Holtorf, Macdonald and Penrose2020). By naming risk and promising a remedy, actors decide what counts as Tuvaluan culture at risk, lift it from its relationships and render it as portable records; they then classify and index those records, place them on external infrastructures, and install regimes of access that determine just who may see them and act upon them. The twin, therefore, accumulates and orders traces under the sign of emergency and binds description to intervention. At the same time, preservation converts the island into an asset twice over. Culturally and morally, it produces value for the rescuers who can point to “our technology” preserving “their heritage.” Computationally and institutionally, it also yields a governable data asset whose value rests in circulation, analysis and control.
4.2 Archive island
In this section, I draw on meeting protocols and interviews with practitioners involved in early-stage digitization to analyse instances in which the twin is clearly an archive, as well as instances in which it functions archivally. I pay attention to how Tuvaluan culture, which has so far been preserved largely through oral tradition and limited national recording practices, is undergoing an expedited process of digitization and datafication. This process aims to produce a digital cultural repository and, more broadly, a machine-readable and operationalized record-keeping entity. It therefore demands close attention to how datasets are collected, organized, distributed and deployed. Critical archival theories of access, consent and custodial power are crucial for considering how the emerging Tuvaluan dataset might be approached responsibly, and for identifying the mechanisms through which communities already made vulnerable can be subjected to further forms of harm (Noble Reference Noble2018; Thylstrup et al. Reference Thylstrup, Agostinho, Ring, D’Ignazio and Veel2021).
When social memory becomes metadata, what is gained in resolution and what has gone beyond retrieval? In a culture shaped primarily by oral tradition, family-based knowledge that was kept as secrets, and ephemeral material culture, the notion of digitization already comes with great concerns. In an interview, the researcher and climate activist Taukiei Kitara noted that the Tuvaluan language has no single word for “repository” or “heritage.” This symbolic anecdote captures the difficulty of archiving a primarily oral culture through digital and automated means, a task that was given to the DAI and the United Nations’ RNI, who work closely with Tuvalu’s Department of Culture. In a roundtable with the organizations’ representatives in Funafuti, Tuvalu, on April 21, 2025, members proposed that sources should be certified and authenticated by the department as genuine data sources, and that access should be restricted to “some of these data sources due to the secrecy or confidentiality and privacy of the data or specific cultural knowledge that belongs to specific family members.” So far, family secrets have been kept through oral traditions, which are often practised but have never been documented through text or sound recording, forms that represent more conventional understandings of record-keeping and provenance.
Importantly, the Tuvalu National Library and Archives in Funafuti mainly holds remnants of colonial correspondence, land titles and genealogy queries, all remains of the period when the islands were under British colonial rule, while elders’ stories remain stored within community memory. Across conversations, knowledge appears as embodied, intergenerational and guarded. Family secrets, skills for healing, fishing and reading the weather are taught through stories and practice. For example, the notion of a national archival entity itself has been challenged by the elderly who insisted on per-island archiving systems rather than having an abstract nation; a 2025 roundtable of community leaders recommended that documentation proceed per island and that land serve as a primary organizing principle, including any rulings from land disputes and courts: “We believe that each island is unique in its own culture, cultural practices, and knowledge. Each island has its own land, history, beliefs and religion. Therefore, we need to document these histories from all the different islands.”Footnote 9 The falekaupule, a traditional assembly of elders which is completely embodied and oral, exemplifies the classification challenge, where the seating, speaking order and the meaning of posts vary by island and thereby encode status and authority. Even a photograph of a canoe, says Kitara, raises questions of classification on whether it should be followed by builders, clan rights and relational obligations, indicating that the archive’s categories will perform culture even as they preserve it.
These discussions are complicated in any context. The Tuvalu case, however, provides an especially acute example for several reasons. Firstly, because self-documentation has never been central to their tradition, Tuvaluans must rely on rescue and preservation methods developed elsewhere, many of which remain Eurocentric and continue the grammar of rescue discussed earlier. Secondly, if the repository is designed for community use, access depends on forms of digital literacy that are often generationally biased, leaving older Tuvaluans at risk of exclusion. Thirdly, if the digital object and evolving dataset are not made only for preservation, but also become actionable in the context of a landless state and a digital twin, then objects once used, felt and embodied become objects known through viewing and simulation. Fourthly, while data ownership may be clarified within the cultural repository itself, where the community is intended to retain full ownership, it remains unclear where data will be stored and who will own or govern it within the broader ambition of a landless, datafied Tuvalu. Without attention to ownership, jurisdiction and infrastructural control,Footnote 10 digitization risks extending archival harm rather than repairing it.
Kitara’s dilemma is echoed in Morehshin Allahyari’s assertion that every act of digital preservation is constitutive, a process of “re-figuring.” As she notes, “It’s an act of going back and retrieving the forgotten or destroyed histories” (Allahyari and Paul Reference Allahyari and Paul2019, p. 2). As Evans (Reference Evans2025) warns, the danger lies not only in the loss of land to climate change but also in the erosion of depth and complexity when cultural artefacts are reduced to commodities within markets for digital experience.
Above, I analysed the cultural repository as a clear example of a digital archive. I now turn to state-scale collective record-keeping and examine how it functions archivally. The choice to use a digital twin for nation preservation is not an obvious one. A twin assembles structured and unstructured sources into a multimodal “live copy.” The act of twinning binds description to action, so records do not sit inert, but rather enact anticipatory decisions and governance (Amoore Reference Amoore2025). The act datafies and fragments entities into re-assembled parts (Korenhof et al. Reference Korenhof, Blok and Kloppenburg2021) and materializes potential behaviours, turning whatever it touches into a data commodity (Ulrich Reference Ulrich2025). In this case, preservation and state digitization are folded together: Tuvalu’s twin is figured at once as an archival entity for record-keeping and cultural preservation and also as a governing entity that will act in and through a digital domain. This twin reallocates archival authority between Tuvalu’s in situ institutions and any ex situ computational infrastructures.
Thus, choosing the twin as the main archival entity for knowledge preservation means that 3D simulation and AI techniques become embedded in archival record-keeping. This places additional pressure on core archival concepts such as provenance, traceability and transparency. Tuvalu’s physical environment, cultural value and heritage must be translated into machine-readable forms that can be authenticated, verified and reused. These requirements are shaped by markets and institutions that reconfigure authenticity, authorship and ownership (Colavizza et al. Reference Colavizza, Blanke, Jeurgens and Noordegraaf2022; Thibodeau Reference Thibodeau and Lemieux2016). As Ribes (Reference Ribes2017) notes, while some memory processes remain documented in the metadata that travels with the dataset, many others “are left in archival traces and memories” (p. 1518) and for others no documentation remains at all. Since these datasets rely on distributed collections created and reused by a wide range of actors across new infrastructures and platforms, they raise difficult questions about how to engage with datasets critically without inflicting second-order violence (Thylstrup et al. Reference Thylstrup, Agostinho, Ring, D’Ignazio and Veel2021). In summary, then, Tuvalu’s archival entity is not primarily one that archives the nation and its knowledges, but rather keeps a record of emerging infrastructures that are driven by a sociotechnical imaginary.
4.3 Proxy island
Most writing on the ex situ scenario since its articulation (Burkett Reference Burkett2011) comes from the law (Canavan Reference Canavan2019), climate policy (Bessette Reference Bessette2025; Comar Reference Comar2024; Jalil Reference Jalil2025) and international affairs (Rothe et al. Reference Rothe, Boas, Farbotko and Kitara2024). This writing largely frames “virtual nationhood” as inevitable, with corresponding shifts in recognition, law and governance. Drawing on critical technology studies and archival discourse, this section proposes that an ex situ digital nation depends on a dense ecology of proxies – human and nonhuman – that enables the production of knowledge through maintenance, authentication and circulation. These proxies include people and organizations (ministries, NGOs and vendors), as well as code, datasets, networks, interfaces, data centres and storage solutions. Taken together, they enable and surrogate the nation at a distance and mediate its cultural knowledges. When record-keeping and governance are routed through the twin, the ex situ island nation appears as a proxy island: A techno-cultural assemblage shaped by subnational, national and supranational authorities, and by computational and global corporate infrastructures. This evolving constellation not only exceeds archival logic but also clarifies how management frameworks and governance arrangements for the island are shaped by external representations that group and separate knowledge according to specific interests. These partnerships often claim superior agency over local social and environmental futures, thereby becoming an emerging manifestation of knowledge islandization.
The realization of Tuvalu as a digital ex situ nation demands a major infrastructural upgrade for a nation that has until very recently had almost no Internet connection and does not yet use credit cards. The implantation of Tuvalu’s Vaka Cable, the collaboration between Google and the Tuvalu Telecommunications Corporation branch of the Bulikula cable system, marked Tuvalu’s first international subsea connection.Footnote 11 In addition, the Falepili Union, which is crucial for any ex situ scenario, turns Tuvalu’s loss into a regional asset by tying climate assistance and mobility pathways to Australia’s wider Pacific security architecture, including influence over Tuvalu’s future security arrangements (Kitara and Farbotko Reference Kitara and Farbotko2023). Next to the above examples that are already implanted, the realization of any Tuvalu twin involves a network of corporations, nonprofits and international partners, all governing “soft forms,” including collaborations between governments, companies, multinational umbrella institutions and international organizations, as well as “proxy organizations” and trade forces. Next to the human actors, the twin solution relies on nonhuman actors, such as training datasets and labels, model weights, Large Language Models (LLM)s for transcription and description, computer-vision pipelines for 3D capture, simulation models and cloud hosting, each of which embeds decisions about measurement, classification, interface, processing power and system boundaries, thereby forming a new political mesh.Footnote 12 In other words, preservation becomes decentralized through infrastructural fragmentation, dividing the nation across the very systems assembled to secure its continuity.
Custodial authority is thus redistributed across institutions and infrastructures that claim the right to describe and interpret. These archons shape the politics of memory by guarding and interpreting the archive and thereby controlling access to legal and administrative records. Their very position grants them authority over what is remembered, what is preserved and how it is interpreted. boyd and Crawford’s notion of “algorithmic archons” (boyd and Crawford Reference boyd and Crawford2012) indeed argues that those who control data analytics wield significant authority over knowledge production, effectively acting as modern-day archons who guard and interpret archives, and shape collective memory and knowledge. By drawing this parallel, boyd and Crawford highlight the performative nature of data analytics, where the act of analysing data is not a neutral process, but one that actively constructs and influences societal narratives and power structures. Keeping in mind the multiple agencies and governing bodies that assert their power and interest on the becoming of this twin, I argue that Tuvalu’s digital twin presents an emerging formation of the archival archons that rely on humans.
Routing record-keeping and governance through the twin, an apparatus of simulation and intervention, creates new dependencies. The ex situ scenario, which has never been realized yet, risks new tendencies that don’t only rely on regimes of recognition and international law but also rely on private computational environments. In this arrangement, ex situ sovereignty becomes deeply dependent on computational and global corporate infrastructures that shape heritage, national record-keeping, governance and knowledge production. This emerging nation preservation model of proxy sovereignty, where external entities exert control or influence over a nation’s affairs and its shape, will determine what will be recognized as the Tuvaluan identity, knowledge and culture in the future.
Tuvalu has a history of producing new variants of sovereignty and dependence. Initially untouched by the digital era, the nation discovered the potential of its .tv domain when an American company had already claimed it as digital terra nullius, supposedly ownerless territory. After reclaiming ownership, Tuvalu capitalized on it, granting exclusive rights to a foreign company in exchange for much-needed funding. The former Prime Minister, Bikenibeu Paeniu, described this agreement as “the best deal Tuvalu, in our short history, has ever reached,” but also acknowledged that “We’ve been dependent on aid for so long – sometimes our sovereignty has been compromised” (Raskin Reference Raskin1998).Footnote 13 This was not that nation’s first venture into leveraging telecommunications – Tuvalu had previously sold unused phone numbers to offshore customers. Despite navigating global digital markets with caution, the nation ultimately secured a lifeline from using video streaming, with Twitch.tv bringing in an estimated $10 million annually, around 10 per cent of the nation’s Gross National Product (GNP).
In Tuvalu’s current National ICT Policy (drafted 2021; approved October 5, 2023), digital transformation is to be delivered through private service providers, and regional and international bodies, together with computational systems (such as nationwide broadband, extension of mobile services to the eight outer islands, a planned submarine fiberoptic cable, local data centres, cloud services and the adoption of Internet of things). The state also commits to new legal and regulatory frameworks, universal access, cybersecurity, values-based capacity building, and monitoring and evaluation. These arrangements produce tensions that clearly matter for an ex situ nation. Universal access and cultural goals rely on external firms and standards; citizen participation and local content are mediated by platform classifications and metrics; per-island stewardship sits uneasily with national integration; and sovereignty is coupled to infrastructure and cloud dependencies. These proxies set schemas and model parameters, and governance institutions then run through computational environments as much as through recognition and law.
An ex situ digital nation depends on a dense ecology of proxies, both human and nonhuman, for the maintenance, authentication and circulation of knowledge. These proxies function as emerging forms of archons (Derrida Reference Derrida2015), combining human and nonhuman dependencies and operating across scales from institutional to national and subnational, and across both the public and private sectors. This circumstance means that ex situ cannot be understood only through regimes of recognition and legal dependency; it must also be read through use of the technological and market dependencies that organize proxy governance. The twin solution for the landless nation risks rendering the island into a market asset and secures first-mover advantages, setting a template that travels to other sites and vendors, extends cable networks and data centres, and produces a governable data asset whose access, storage and analysis are controlled by both public and private actors.
4.4 Data island
Based on public sources, Tuvalu’s “digital nation” draws on four main data arenas. First is geospatial capture, where airborne LiDAR of the atolls and surrounding reefs establishes accurate ground and seafloor elevation, funded through UNDP and the Green Climate Fund and implemented with Fugro (Taishi and Webb Reference Yusuke and Webb2024). Public materials add a comprehensive 3D LiDAR scan of all 124 islands and islets. Second is high-resolution imagery: PLACE (a nonprofit data trust), working with the Tuvalu Department of Lands and Surveys and the Pacific Community (SPC), collects aerial and street-level data from Funafuti for mapping assets, such as trees, houses, infrastructure and cultural sites. Third is an environmental time series for planning tools, including the national Hazard and Risk Dashboard that provides wave impact modelling and shoreline change monitoring across all nine islands. Finally, there are the narrative and planning models. Accenture Song produced an Unreal Engine model of Te Afualiku for public communication and immersive viewing, consistent with wider accounts of the twin used for analysis and incident navigation. Capture is delivered by a consortium. UNDP and Fugro provide the LiDAR baseline; SPC and PLACE collect the aerial and street view imagery; and Accenture Song, with its creative partners The Monkeys and Collider, led the initial “First Digital Nation” visual assets and public launch. Data transport relies on national connectivity upgrades, while storage and computation are framed in the National ICT Policy, which anticipates data centres, cloud computing, and the use of machine learning and AI for public services. Specific record-level selection criteria, access tiers, licensing and platform providers for long-term custody have not yet been detailed in the official public documents (Srinivasan Reference Srinivasan2025).
Although the digitization of Tuvalu is done through careful collaboration, not everything can be datafied. Although “high fidelity” and “ultra-realistic” are terms used to advertise twin products, the idea of creating an accurate picture is misleading in practice. Based on interviews with system engineers who explain that capturing focuses only on the specific features that matter, Louise Amoore argues that the twin is contrary to the idea of an accurate copy or replica. “What the digital twin actually ‘wants,’ then, is sufficient precision to make decisions and actions possible” (Amoore Reference Amoore2025, p. 4552). It brings attention to some features of the world and not to others, as only what can be expressed as numerical, visual, spatial, textual or environmental data can be incorporated into the twin. This is a process of “ripping reality,” as Hito Steyerl suggests (Steyerl Reference Steyerl2017), underscoring that the act of capturing depends precisely on these systematic gaps in understanding. The very holes, seams and repairs that make virtual environments stable, both technically and rhetorically, are where power and fiction enter the record.
Drawing out the projected implications from Amoore’s study, in Tuvalu’s case, these gaps matter because the twin is not only a preservation device but also the infrastructural basis for an automated nation. The datafication of land, reefs, houses, cultural sites, risk models and immersive environments prepares Tuvalu to be queried, simulated and acted upon through computational systems. Preservation therefore begins to work through disintegration, breaking lifeworlds into searchable, measurable and actionable units before they can circulate as data. In this sense, the making of the digital twin is not the same as making a usual replica or copy, even though it is often done through a performative appeal to digital copies or replicas that the digital twin is representing (Amoore Reference Amoore2025). This kind of performativity is different from that of the modern archive, which is based on inclusion and exclusion, indexing and mediating institutional knowledge. It offers highly exclusive cyclical digital modelling, and the simulation becomes indistinguishable from the real.
As mentioned before, by positioning digital twins as archival entities that structure the world through a feedback loop, the notion of fidelity can be reexamined. Focusing on data underscores that such fidelity does not operate in relation to the original, nor does it end where Korenhof et al. locate the mechanisms of steering power. Tuvalu’s twin is self-enclosed, but a process of re-entering and reconfiguring the relations between technical configurations and physical realities, based on prior technical, historical and ontological conditions (Hui Reference Hui2019), shapes its sovereignty in an ongoing process of self-modification. As Amoore puts it very clearly, the paradigm of the digital twin is absolutely at the heart of the dreams of automation, where the most important value of the proposed algorithm is to break through the data boundary that exists between the real and the virtual.
For Tuvalu, this dream of automation means that the nation is not only preserved as a record, but prepared as a system that can be simulated and acted upon. The automated nation is therefore not a fully autonomous state, but a state whose cultural, environmental and administrative traces are formatted for computational response. Taking that distinction into account, even when carefully managed, there is potential asset value being produced by the aggregation and discoverability of heritage, and by a geospatial model that can be queried for planning and public communication, which is consistent with the broader accounts of digital twins in public media. Therefore, the question of fidelity should not be addressed in relation to a specific document, or a nation, but to the infrastructure that thrives on crisis.
5. Conclusion: a sea of islands
I’ve read Tuvalu as an emerging digital twin next to archival and performance theory. By doing so, I unpack how being archived carries new and unforeseeable risks and invites a reconsideration of the archontic apparatus of cultural memory in the age of AI and climate crisis. In describing such an ex situ scenario, I use the term “islandization of knowledge” to criticize a colonial technique of extraction and assetization that has now been further developed through a comparison of archival practices and digital twinning.
Against this scenario, I have shared protocols from roundtable discussions with community leaders that indicate a preference for per-island repositories rather than a single national vault, which would be developed island by island so each passes through the appropriate cultural systems and protocols and receives the blessing of its leaders and decision-makers. These two opposite perceptions were captured in the words of Epeli Hau’ofa, who proposed a shift from viewing the Pacific as “islands in a far sea” to “a sea of islands” (1994, p. 189). The first view isolates dry landmasses at the margins of power, while the second concept foregrounds circulation and relation, where that actual scale arises from networks of relationships rather than the distance from centres.
Beyond legal and diplomatic framing, ex situ nationhood is a technological formation. The platform selected for preservation, the digital twin, is not a neutral repository but a structuring apparatus that turns social memory into metadata and binds description to action. In the case of Tuvalu, the technical and institutional stack that includes public agencies, NGOs, private vendors, code, datasets, networks, data centres and storage systems configure what is recorded, who can act on it and to whose benefit. Treating the twin is/as archive clarifies how this sociotechnical imaginary will shape heritage, provenance, record-keeping and authority in the ex situ arrangement.
In the era of deep learning and generative AI, the digital twin increasingly operates within a model-driven paradigm of pattern discovery across heterogeneous and loosely structured traces. This shifts archival practice from description towards prediction and generation and links classification to decision and resource allocation. Although the twin is often presented through a performative claim to faithful copying, it moves away from a mimetic into performative logic. By correlating unstructured data streams and recombining them, it can compose people, knowledges and memories and model new interactions between submerging islands and their virtualized ex situ state.
Foregrounding possibilities to reimagine both the content of digitized collections as well as the conditions in which we encounter them, I join Echavarria and Rothe et al. in underscoring the digital Tuvalu project’s potential to act as a merger of Indigenous cosmologies with Western institutions through a networked relationship between physical and virtual existence. This calls for a critical and reparative archival perspective that offers the possibility of imagining a future different from the present. A micro-local, per-island archive aligns with an Indigenous imaginary that treats islands as nodes in a moving sea, each sustaining a situated culture. Tuvaluan wayfinding, grounded in relationships rather than fixed locations, offers a method for knowledge transmission that resists archival fixities. To imagine the archive as a sea of islands, with Tuvalu’s elders, is to imagine a future-oriented practice that situates local knowledge within uncertain technical and political conditions and allows learning from mistakes rather than simply codifying them and insisting on locality, relationality and non-scalability as the only thinking principle for an incomplete future.
Acknowledgements
The author thanks Dr Annette Kühlem, Head Office Project Lead of the Tuvalu Digital Repository for Cultural Heritage, and Dr Taukiei Kitara, Griffith University, for their generous support and insights. She also thanks Dr Noa Roei for her attentive reading of the final draft, and Dr Roei and Dr Annet Dekker for their shared work on digital twin imaginaries as part of the Centre for the Study of the Networked Image at the University of Amsterdam.
Author contributions
The author was solely responsible for the research, writing and revision of this article.
Ethics and approvals
This article does not contain any studies involving human participants performed by any of the authors.
Funding statement
The authors acknowledge TU Wien Bibliothek for financial support through its Open Access Funding Programme.
Competing interests
None.
Dr Ofri Cnaani is an artist and researcher who works in performance and media. She is a Guest Professor at the Institute of Visual Culture, TU Wien, and a research fellow at the Amsterdam School for Cultural Analysis, University of Amsterdam. Her work examines digital heritage, data coloniality, performance and critical technology.