Algorithmic Heritage and AI-Assisted Museums in Morocco and Egypt—From Clouded Coloniality to Techno-Cultural Empowerment

Abstract

This article provides a comparative examination of two AI-assisted museums in Africa, the Dar Gnawa Museum (Marrakech) and the Grand Egyptian Museum (Cairo). It analyzes the AI functions and the strategies these institutions adopt to pursue techno-cultural empowerment in a field long shaped by power asymmetries in Africa. The literature review highlights how technological transfers in museum cooperation remain an overlooked vector of coloniality which the convergence of AI and heritage practices now brings sharply into view. This article develops the notion of clouded coloniality—a dual phenomenon in which heritage data is literally managed in the cloud, often from abroad, while diffracted layers of actors and processes obscure the identification of new imbalances in and around AI-assisted museums. This article designs a two-pronged analytical framework which first assesses AI functions within the Dar Gnawa Museum and the GEM, and then evaluates sustainable synergies between these institutions and their broader AI ecosystem. The results indicate that whereas the GEM prioritizes youth empowerment, the tourism industry, and partnerships with foreign stakeholders that could potentially expose Egypt’s cultural sovereignty, the Dar Gnawa Museum independently developed an AI tool, Kouyou, that could offer a transferable model for advancing Pan-African techno-cultural empowerment.

1. Introduction

Though intuited by Alan Turing [1] and academically introduced at the Dartmouth conference in 1956, it is only over the past few years that data-fed AI has become the driver of the Fourth Industrial Revolution. “Data is the new currency and access to data—rather than money, natural resources, or advanced weaponry—is now the most valuable asset available to nation-states and corporations” [2]. The advent of AI has had sweeping implications in every sector, including museums, from now on characterized as “socio-technical ensembles that constitute, stabilize, and transform the constantly changing relations between AI technologies (computer vision, natural language processing, artificial neural networks, et cetera), human beings (museum staff, researchers, IT experts, visitors, users, artists), material objects (historical artefacts, artworks), and real or virtual environments (exhibition spaces, digital archives)” [3]. These dynamics raise ethical concerns regarding transparency, copyright, algorithm bias, privacy, human oversight and data ownership [4]; especially so in Africa where museums, once colonial institutions, have continued after independence to rely on Northern actors for financial aid, technological support, and professional expertise [5].

A cursory glance at the cartography of AI in Africa unveils conspicuous paradoxes even to the untrained eye. Africa is far from “lagging behind”, as is often assumed in development or museum cooperation literature. It currently holds five museums, well distributed across North, West, East and South Africa, primarily dedicated to or relying on AI—Iziko Museums’ Planetarium and Digital Dome (Cape Town, 2017); Ethiopia’s Science Museum (Addis Ababa, 2022); the Grand Egyptian Museum (Cairo, 2024); Dar Gnawa Museum (Marrakech, 2024); and the International Museum of the Prophet’s Biography and Islamic Civilization (Dakar, 2025). Contrary to all previous phases of technological transfers in museum cooperation, none of these five leading institutions relies on direct Eurocentric North–South cooperation. Ethiopia cooperates with China; Senegal with Saudi Arabia; South Africa relies on its own AI capacities; Morocco and Egypt—who obtained loans and capacity transfers from Japan—even portray themselves as technological leaders in Africa [6,7]. However, a more disturbing reality emerges when looking at the distribution of AI data centers across the world. While North America and Europe, respectively, host 41% and 23% of them, Africa only has 233 data centers—56 of them in South Africa alone—spread across 38 countries, that is, 2% of the world’s total [8].

This raises the familiarly uncomfortable question of who ultimately stores and owns African heritage and knowledge, this time encoded in AI software, hardware and infrastructures. Underlying this topical concern lies a vacuum in the literature whereby scholars and practitioners have underestimated the extent to which technology and capacity transfers in museum cooperation with Africa—from the dioramas in the 1930s to most recent digital initiatives since the 2010s—constituted an insidious lever of coloniality within institutions originally imported from the North [9]. As a result, the hypothesis whereby technologies could be a catalyst for (de)coloniality within cultural institutions such as museums has been obliterated. But in the present context where a growing number of African institutions resort to AI—in addition to the abovementioned museums, the Museum Africa (Johannesburg), the Egyptian Museum (Cairo) and the National Museum of Egyptian Civilization (Fustat), either hosted temporary exhibitions or have ongoing AI programs—bridging this conceptual terra incognita between the inherent colonialities of AI [10,11] and museums [12,13] seems necessary, especially at a time when new South–South and intra-African diplomatic allegiances are reshaping the continent’s museumscape.

This paper offers comparative perspectives on the Dar Gnawa Museum (Morocco) and the GEM (Egypt). The Dar Gnawa Museum emerged in Marrakech’s Derb Jdid district during the 1960s, alongside the establishment of another Dar Gnawa center in Tangier. It traces the history of how enslaved sub-Saharan communities settled in the Maghreb since the late 19th century. While its collections include guembris, ritual attire, and a variety of artifacts that reflect both the continuities and the transformations of Gnawa society over time, its main focus is on music, traditional ceremonies and other forms of intangible heritage. The museum extends its reach beyond Marrakech by showcasing pieces at international events, such as the Dakar Biennale held in November 2024 [6,14]. Construction of the Grand Egyptian Museum began in 2005. Although sections of the complex were opened to the public in 2024, the official inauguration took place in November 2025. The museum occupies a site in Giza, approximately two kilometers from the Pyramids and the Sphinx, thereby forging a deliberate symbolic connection between Egypt’s ancient monuments and its most ambitious contemporary cultural project [15]. Covering almost 480,000 square meters, the GEM is designed to be the world’s largest archaeological museum. Its holdings exceed 100,000 objects—including the entire funerary assemblage of Tutankhamun—and trace the span of Egypt’s Pharaonic civilization. Both institutions emerged in countries that suffered from colonization. Between 1912 and 1956, Morocco was under French and Spanish rule, while Egypt was occupied by Britain from 1882 to 1956. Colonial-era acquisitions, excavations, and trade dispersed their heritage across Northern museums [16,17]. After independence, Morocco’s Gnawa culture remained marginal until its revalorization in the 1990s [18]. In Egypt, museum policies alternated between the nationalization of the 1950s, restitution demands and, conversely, tourism-oriented expansion and ODA-financed mega-projects promoting heritage branding and soft power [15]. Today, both countries’ museumscapes continue to reflect “a local history of colonial museology and, at the same time, (…) political ambivalence over the role the museum should play in the nation’s postcolonial life” [19].

Based on these two case studies, this paper addresses a double-layered research question. At a curatorial level, it explores the extent to which these institutions have integrated AI functionalities and whether these changes reshape and decolonize existing curatorial practices. While necessary to understand these institutions, this technical perspective yields incomplete answers to the core question of our investigation, that is, whether the Dar Gnawa Museum and the GEM suffer from enduring or renewed forms of coloniality. Indeed, coloniality, briefly defined as a persistent system of domination that may subsist after the historical period of colonization ended [20,21,22], can be polymorphous and invisible, all the more so in its AI variant often portrayed as neutral [23]. In a second phase, our investigation thus examines how the Dar Gnawa Museum and the GEM respond to what we term clouded coloniality—a term that seeks to capture both the elusive nature of coloniality and the fact that heritage data is now literally hosted on cloud infrastructures. The discussion focuses on six dimensions of techno-cultural empowerment: legal frameworks and infrastructures, social inclusivity, STEM education, innovation capacity, private partnerships, and AI diplomacy. After setting out a comparative discussion, this paper ends with concluding remarks on the need to bring together the technological and cultural facets of museum cooperation so that AI becomes an empowering tool for African institutions.

2. Literature Review

2.1. Technological Transfers to African Museums, a Vacuum in the Literature

Science and African museums have always had a tumultuous relationship. While local customs on conservation of artefacts had long existed [24], Northern museums were exported to Africa during the colonial era everywhere except in Gabon, Equatorial Guinea and Djibouti. They were used as “vehicles to perpetuate colonialism that strategically otherized indigenous people and their cultures (…) seen as barbaric, backward, evil and uncivilized” [19]. Similarly to schools and churches, which deliberately proselytized, museums imposed epistemic categories (ethnography, anthropology) to objects (literally artefacts but also ethnic groups with their traditions and rituals) they interpreted by silencing and stigmatizing local voices and knowledge [25,26]. While they casted African tools as obsolete, collections were exhibited in architecturally outstanding buildings—the colonial staff’s houses, reconverted post-offices and other administrative buildings—supposed to embody an architectural and technological strand of modernity that further contrasted with local artefacts and architecture [27]. However, museums’ scientific message had limited radiance as they would only open their doors to the general public after independence [28]. From that time on, the bulk of equipment and technological transfers would consist of relatively modest and often outdated tools, such as vitrines, lighting systems, audio guides, conservation materials, and photographic equipment, with exchanges limited to photographic archives, cataloguing systems, and preservation protocols. Moreover, the budget earmarked for technology within museum cooperation remained almost untraceable—a list of interior decoration items or IT equipment is sometimes laconically included in some museum cooperation blueprints [29,30,31,32]. These reflected an asymmetrical relation whereby the recipient’s requests were reformulated according to the donor’s will [29].

Compounding this predicament, transfers and trainings on the use of technological devices within museum cooperation lie at the intersection of a double vacuum in the ODA literature. On the one hand, cultural—and therefore museum—projects do not constitute an acknowledged category in the OECD taxonomy. They are instead diluted into the blurred margins of social infrastructure and multi-sector/cross-cutting programs [33]. On the other hand, STEM capacity and equipment transfers are equally obscured within education or economic infrastructure projects [33]. Studies on STEM capacity transfers focus on their impact on FDI [34] or on the donor’s exports [35,36], but not on the cultural empowerment deriving from technologies. Therefore, there is a need for critically engaging that ahistorical technological inheritance in the heritage field [22]. Besides, until the opening of Ethiopia’s science museum—the first African museum dedicated to endogenous science—African technological artefacts were relegated to temporary exhibitions and archeological projects—conveying the message that Africa’s science belongs to the past [37,38]. African museums have thus long been caught in a double bind whereby techno-cultural empowerment is eclipsed by higher ODA or curatorial imperatives.

2.2. The Limits of Digitization

Prior to current AI advances, the digitization of collections in museums intersected with decolonizing discourses in a clear-cut manner, with virtual museums, material replicas and digital restitution questioning the borders of possession and materiality in collections [39]. Digitization opened new perspectives on dialogues on heritage but also delivered mixed results. On the one hand, it dangled the promise of extending museums’ contact zones between the institution and the communities it generates knowledge on from the material to the virtual spheres [40,41]. Digitization, it was hoped, would bridge the physical and intangible gaps generated by socio-spatial estrangement and reverse exclusionary path-dependency [22]. Workshops in Zambia, for instance, “digitally repatriated” objects in the diaspora to communities of origin with a view to restore epistemic gaps. As it turned out, communities were not only able to confirm or deepen the information associated with certain items, but in some cases, they were also able to re-learn knowledge systems that were lost through inculturation [42]. Ultimately, this information was shared on a digital platform, like many others such as Digital Benin or Le Monde en Musée.

Investigating the issue of ownership beneath the materiality of objects therefore leads to the question of who runs the digital platform and ultimately controls access to the data encoded as cultural capital. In Gabon, for instance, President Bongo decided to launch Le Musée Virtuel des Arts et Traditions du Gabon in 2006. With 85% of the online pieces coming from European museums, this project furthered the decorrelation between places of provenance and places of display as it did not question cultural ownership by the former colonizer [43]. More worryingly, the virtual museum’s technical and epistemic specifications were outsourced to Novacom, a French multimedia agency based in the former metropole [44]. In other words, in addition to retaining heritage, France also refrained from transferring the skills and equipment that could have granted Gabon autonomy in managing its virtual collections. Likewise, the oxymoronic digital restitutions conducted by Belgium to Rwanda consisted of handing out a USB key, that is the device, rather than practical know-how on digitization or material artefacts [45].

In the two countries under study, Morocco and Egypt, heritage digitization is equally limited in scope and uneven in implementation. The “Egypt Digital Heritage Platform” enabled the digitization and public availability of about 11,300 works from the National Library and Archives, but museums have to do more to integrate communities as co-designers, rather than passive sources, of AI tools [46]. In Morocco, despite various programs to digitize archives, artefacts and traditions, both tangible and intangible heritage remain under-recognized and under-promoted, while heritage governance is frequently fragmented [47]. In sum, digital initiatives have been used as a pretext to avoid, not as a step towards, material returns [6]; and digitization did not empower African museums nor source communities, often deprived of agency in this process.

2.3. When Parallels Intersect: The Converging Colonialities of AI and Heritage in Africa

Inasmuch as AI-assisted museums “enhanced by 3D systems and virtual reality offer immersive, mobile experiences that preserve and share cultural heritage while engaging users through interactive exploration of art and information” [48], AI has erroneously been portrayed as the prolongation of digitization initiatives in the museum field. There are, in fact, important qualitative differences. Whereas earlier digital tools followed pre-programmed instructions that required rules designed by humans, machine learning systems independently identify patterns in data and adapt outputs without direct human intervention [49]. Museums using AI do not simply digitize existing content but create adaptive forms of classification and visitor engagement. These processes, especially deep learning, operate through layered models that tend to mask causal reasoning, generating the well-documented “black box” problem [50]. As a result, museum professionals who delegate aspects of curatorial management to AI may find themselves unable to fully justify certain outcomes. In addition, the accuracy of AI systems relies on access to high-quality datasets fed by extensive data infrastructures [51]. This dependency means that the efficacy of AI in museum contexts cannot be disentangled from wider questions of data governance, accessibility, and sovereignty, as we shall see below.

Looking inside the black box, AI can be succinctly defined as a combination of software (machine learning algorithms, models, programming frameworks) whose intangibility paradoxically relies on hardware devices (CPUs, GPUs, TPUs, ASICs, and edge devices) and material infrastructures such as data centers, computers, chips manufacturing and satellite networks [52,53]. Alternatively, AI can be apprehended through the functions it performs, that is, the application of advanced analysis and logic-based techniques, including sensory perceptions, classification of data, automation of tasks, machine-learning—processes hitherto carried out by humans—to interpret events, analyze trends, automate decisions and perform actions [23,54].

Due to its scientific nature and the technicalities it involves, AI is often presented a neutral or objective tool supposedly free of bias [23]. However, a growing body of literature has warned that algorithms are opinions embedded in code whose biases mirror society’s and can be turned into weapons of math destruction [6,10,55]. Hajri [56], for one, links the wording and concept of “Artificial Intelligence” to a covert ideology. While artificial implies that non-natural, man-made machines can substitute and surpass the abilities of man—a belief common to transhumanism, computationalism or even anthropological nihilism—intelligence “was historically linked to theories of racial improvement and eugenics, thus legitimizing racial hierarchies and justifying the enslavement of certain groups” [56]. Beneath the apparently innocent promise of “tech solutionism” hides a convoluted canvass of prejudices. Authors from several disciplines ranging from media studies and computer science to political economy and social science have in fact denounced “platform imperialism” [57], “algorithmic colonization” [10], “data extractivism” [58], “digital colonialism” [59], “data colonialism” [60], or “tech colonialism” [11]. This plethora of concepts self-evidently share a common criticism of colonialism, but the links between the malign intelligences of colonialism and AI need to be more neatly delineated.

Data colonialism and digital colonialism operate on two overlapping yet distinct planes that are often conflated. Both share a colonial metaphor and draw on power asymmetries, yet digital colonialism is more explicitly understood as the historical continuation of colonialism by technological means. It acknowledges a direct lineage [59,60] where Western companies once extracted raw materials during colonization, they have been replaced by Western tech firms exploiting new communication infrastructures rather than transport networks [2]. Just like information was extracted from Africa, to be then treated and commercialized in the West during colonization, data is currently extracted from Africa and developing countries to be treated in the North [11,22]. Data extractivism is doubled by a similarly non-reciprocal, one-way resource extractivism, well-established in the literature on textile and mining, which “refers to the appropriation of value by Northern countries and corporations and the alteration of the source of that value” [59]. To name just one example, the Democratic Republic of Congo accounts for roughly 70% of global cobalt production—an essential mineral for AI hardware. And the growing need for other AI-critical minerals—coltan, lithium, graphite, and copper, inter alia—“is projected to lead to a 3500 per cent rise in lithium demand in the European Union by 2050” [59]. Digital colonialism can thus be framed as a concrete, historically embedded system of oppression whose violent legacy persists through computation [59]. By contrast, data colonialism is not limited to former colonies in the Global South. It may unfold in the Global North as well [61], generating social and spatial injustices beyond marginalized communities through uncontrolled e-governance [62] and pre-structured biases in platforms’ affordances [63]. While both forms involve data extraction, Le Ludec [64] nuances the distinction: Northern Internet users may unknowingly become subjects of data colonialism, while Southern data producers are despite themselves exploited to generate data. The contrast is further deepened by the digital divide, especially regarding access and affordability of data: Internet penetration in Africa averages around 40%, compared to over 90% in the Global North, while the relative cost of data is nearly ten times higher [65]. The divergence becomes clearer when considering control. Digital colonialism highlights ownership and governance of the digital ecosystem—hardware, software, platforms, undersea cables, satellites, AI systems—dominated by states and corporations [66]. Data colonialism, on the other hand, foregrounds the commodification of life into data streams, largely through global corporate platforms such as Google, Meta, Amazon, Huawei and Tencent. In Africa, these forms of interference have fueled debates over global power imbalances induced by loss of digital sovereignty [67,68]. This prompted the African Union to adopt its Digital Transformation Strategy for Africa (2020–2030) [69]. As of 2025, 19 African countries have already implemented national AI strategies [70]. Foreign interferences resonate peculiarly on the continent, where the historical “civilizing mission” of the North finds echoes in today’s “digitizing mission”. Behind the presumably neutral idea of “tech solutionism”, data are shaped, constructed and designed to discriminate [67], so are discourses on data which all promote “progress,” “development,” “doing good” [71], “connecting the unconnected” under the pretext of benevolent altruism, humanitarian ethos and philanthropic contribution [11]. These apparently selfless intentions share uncanny similitudes with the colonial civilizing mission and claim to universality [72]. Representative of European colonialism, for instance, statues in the main hall of the AfricaMuseum (former Belgium Congo Museum) can still be seen as didactic allegories of the so-called “civilizing mission”, bearing inscriptions such as “Belgium bringing well-being to Congo”, “Belgium bringing security to Congo”, and “Belgium bringing civilization to Congo”. But what empirical evidence could support the argument that there are converging parallels between the colonialities of AI and heritage in Africa?

The extractivism of information, values and materials are in fact common to the heritage field. While only 2% of the world’s data center are located in Africa, under 5% of the African heritage is stored in African museums [73]. The similarity of proportions between data centers and African heritage located out of the continent cannot be imputed to a mere coincidence. To understand this parallel, it is necessary to look at the broader structure that governs the lifecycle of data, as analyzed by Montoya and Roberts [60], and that of African artefacts (Figure 1 and Figure 2). In the same way as massive data sets are produced through the labor of poorly paid data annotators—many of them based in Africa—who perform the painstaking task of labeling and classifying training inputs for machine learning models [74], artefacts in Northern institutions were first created by source communities. Both were then “collected”, a euphemism often hiding lack of consent and violence. Today’s AI infrastructures extract data, often without consent or reciprocity, from vulnerable populations [62,75]. In the similar fashion, the language used in museum archival records often served to obscure the realities of looting artefacts, effectively exonerating both the original perpetrators and subsequent holders from accountability [76]. Instead of being first stored and then managed as suggested by Montoya and Roberts [60], we would revert the order of these two steps, and propose that objects were first epistemically analyzed before being managed. That is to say, if considered of high value, they were shipped to Europe where they were sold to either private collectors or public museums; of lesser value, they would stay in the colonies’ museums or in colonial administrators’ own collections [77]. These objects were then attributed aesthetic and financial values, as opposed to the more social and ceremonial uses in source communities. Finally, and again contrary to African customs whereby objects are either hidden, burnt, buried or recycled into other objects [78], museum items are conserved. To a great extent, the lifecycles of data and objects reinforce asymmetries of control, ownership and knowledge. They also highlight that the extra-territoriality of information and collections constitutes a breach on the techno-cultural sovereignty of African nations, hinting at enduring forms of coloniality.

This section hypothesized that given the limits of digitization as a decolonial tool and the so far unnoticed convergence of colonialities of data and heritage, there are risks that the integration of AI into African museums could further accentuate unequal practices and algorithmic dispossession [22]. Indeed, just as dislocated ownership has long characterized African artefacts dispersed between mostly Northern collections and African museums under the ambiguous notion of “shared heritage” [79], so too does algorithmic heritage—defined as the translation of cultural practices and artefacts into digital formats that are governed by data-driven systems—presents features of an emerging strand of coloniality whose boundaries, being indistinct, convey a clouded picture for various reasons. Whereas colonization was easily identifiable with its pageantry of conquests, flags and blood, one of the key characteristics of coloniality, all the more so AI-generated coloniality, is precisely its elusiveness. Coloniality is often muted or invisible [80]. By way of illustration, our Ethiopian interlocutors did not acknowledge it, contrary to our Moroccan and Egyptian interlocutors at the Dar Gnawa Museum and French University in Egypt (UFE) who stressed the paradoxical inevitability of exposing their cultural sovereignty to foreign interference in order to exchange and upgrade skills and expertise [6,7,46,81,82]. This coloniality is further clouded by the fact that, as with data colonialism [63], it does not necessarily emanate from Northern governments or governments at all, but instead from a transversality of fields and actors [2], at the center of which software, hardware, and data-brokerage companies play a key role. Finally, algorithmic heritage is being literally clouded to data centers whose localization, management and ownership lie beyond the scope of the originating communities or institutions. As we have sought to demonstrate, the implication of this for African museums is that technological and cultural cooperation can no longer be apprehended separately as technological coloniality may conceal itself behind the promise of cultural empowerment; or, put differently, a new form of clouded coloniality may surreptitiously erode African museums’ efforts toward genuine techno-cultural empowerment. Before further refining the concept of clouded coloniality and its antonym, techno-cultural empowerment, it is now necessary to turn to the materials and methodology employed to probe the diffracted data and stakeholders in the museumscape.

3. Materials and Methodology

The new definition of museums approved by the International Council of Museums (ICOM) in 2022 makes no reference to the use of digital technologies [83]. This absence is all the more conspicuous given that a growing number of institutions are either directly dedicated to the NICTs or increasingly rely on them, especially AI. As astutely noted by de Mello et al. [54], turning a blind eye to technologies in museums will “not make them disappear, but rather deepens the gap between different institutional realities and institutional and geographical asymmetries”. Museums have long embraced technological innovation, from the panopticon (1791), panorama (1792), and diorama (1823), which introduced new modes of interactivity [84], to the consistent integration of contemporary advances—from industrial machinery and space exploration to wartime technologies, telecommunications, and early digital computing [85]. The latest wave of innovations integrating the museum microcosm is naturally dominated by the advent of AI. Smart museums based on AI technologies now “animate otherwise static collections, creating an experience that connects people, education, culture, and art” [4].

Though ultimately focused on the Dar Gnawa Museum and the GEM, our methodology counterintuitively takes non-African institutions as starting points or rather as starting counterpoints. The reason for this preliminary step is simple: if African museums were not to deploy the whole spectrum of functions other AI-enabled museums deploy elsewhere, this gap could be interpreted as a technological lag propitious to asymmetrical relations in potential capacity transfers—either through ODA or museum-to-museum cooperation—identified above as markers of coloniality. On the contrary, should this hypothesis proved unfounded or should African museums partially address coloniality within conventional curatorial functions, it would signify that other benchmarks are necessary for the purpose of assessing a potentially (neo)colonial presence driven by AI. The aim here is not to impose Northern prisms on African museums, but rather to establish a preliminary taxonomy of AI functionalities across museums worldwide and detect potential divergences between the Dar Gnawa Museum and the GEM, and non-African institutions.

Drawing on publicly available museum documents, websites, and exhibition descriptions [Appendix A], it emerges that a first category of AI applications revolves around inventorying, archiving, cataloguing and, in general, managing information related to collections [23]. Indeed, inventorying collections within museums seemed to be predestined for automated indexing processes [85]. By way of illustration, the Harvard Art Museum [86] applied computer vision and natural language processing to generate over 66 million machine-generated descriptions and tags. However, transformative technologies extend well beyond operational efficiency [4]. Still related to collections but for research purposes, a second category relates to how machine learning and computer vision are being used to analyze “deep features” and virtually reconstruct original artworks. By automating the identification of objects, people, clothing, landscapes, and iconographic motifs in paintings for instance, scholars may examine an artist’s line work with unprecedented granularity. At the Rijksmuseum [87], neural networks now quantify stylistic variations, resolve attribution disputes, and support conservation through digital reconstruction. Here, AI-driven automation helps advance research by revealing previously unexplored dimensions of collections.

Moving from collection-centric to visitor-centric functions, machine learning for visitor data analysis creates customer predictions based on previous visitors’ behavior or conducts a sentiment analysis of a museum’s visitors. Various digital touchpoints, which integrate cloud-based analytics through software—like the MoMa cloud-based CRM platform powered by Salesforce [88]—gather pre-, during, and post-visit data not only to dissect preferences and predict trends [89,90], but also to tailor narratives, content, and interfaces for future exhibits [4]. In this vein, “museums evolve from being collection-based and information-oriented repositories of knowledge, with visitors as passive recipients, to adopting participatory approaches that position the audience as central to experiences, making them active participants in value co-creation” [89]. This expanding field has been labeled edutainment, straddling, as its name suggests, the borders between educating audiences and captivating them in the ever more competitive landscape of the entertainment industry. Examples of edutainment devices may include immersive rooms (Museum of the Future, Dubai) [91], holograms (Illinois Holocaust Museum & Education Center, Skokie, IL, USA) [92], interactive screens, objects and murals (The World of Leonardo da Vinci, Milan, Italy) [93], 3-D models (Smithsonian Institution, Washington, DC, USA) [94], Virtual Reality (VR) and Augmented Reality (AR) devices (DuSable Museum, Chicago, IL, USA) [95]. In this respect, gamification plays an essential role here as it increases “engagement with cultural content through affective triggers, and [creates] immersive and memorable experiences that improve users’ learning capabilities” [4].

Mixed reality (MR) produces a new phygital environment where the real and virtual tend to form a continuum [96]. Here, artistic creation with the assistance of AI emerges within the fluctuating frontier and interactions between real and virtual environments. Indeed, such works often merge physical materials and algorithmic generation, thereby creating hybrid spaces in which the analogue and the digital are constantly negotiated. AI provides new avenues by assisting artists in co-creating, thus “positioning technology as a catalyst for human creativity and collaboration rather than merely a tool for efficiency” [89]. In this line, “Data Dreams” or the Karlsruhe Center for Art and Media (Germany) seek to actively reshape the environment by addressing pressing issues such as how algorithms and datasets challenge our perception of reality; the immense environmental costs of the data-driven economy; and the co-evolution of humans and machines [97,98]. This intertwining between oneiric and material dimensions, long practiced by Aboriginal Australians, is now re-explored at the Museum of Contemporary Art Australia where artists directly engage with AI technologies. In this vein, Seoul Robot and AI Museum [99] also tries to fill the void between physicality and the virtual with various media as artistic materials. The media artist group TeamVOID designs experimental systems through kinetic sculpture, light sculpture, and robots to explore the fusion of technology and art.

These reflections highlight the fact that museums are not microcosms in and on themselves [100]. While they seek to capture the world, they remain connected to their external environment. As underlined by the ICOM [83], museums are “open to the public, accessible and inclusive”; and operate (…) with the participation of communities”. Whether through ethical concerns reflected by society, financial entanglements with their sponsors, political messages they may convey or broader missions as cultural mediators, museums are porous institutions. It would therefore seem justified to add a final category consisting of the extensions museums weave with the outside world beyond their core curatorial, educative and artistic functions. These extensions are understood here as commercial applications, diplomatic initiatives and all other activities that consolidates a museum’s agency in its cross-sectoral collaboration [101]. To develop market opportunities, the RAIM [102] is collaborating for R&D and commercial AI applications with, inter alia, Korea Telecom, Korea Institute for Science and Technology and Hyundai Robotics. At a public diplomacy level, the Smithsonian Institution [103] leverages AI—through partnerships with Amazon Web Services and Microsoft—to digitize and translate vast collections, positioning itself as a global digital heritage hub that advances U.S. cultural diplomacy through the projection of soft power. These instances suggest that AI, even from a museum perspective, has to be understood as a tentacular tool with extensions far beyond the museum precincts.

This brief overview indicates that museums resort to AI for the following purposes: inventorying; research; visitor management; edutainment; artistic creation; and other extensions, whether diplomatic, commercial or others. In order to evaluate the extent to which African AI-assisted museums resort to these AI functionalities and whether these reshape and decolonize conventional curatorial practices, it may be visually and analytically more helpful to project these functions on a radar chart whose radiuses work as benchmarks scoring from 0 to 5, detailed in

Table 1. Criteria for evaluating AI functions in museums.

Score	Criteria
0	No AI tools are used or considered for this function
1	Awareness of AI exists, but no concrete implementation has been initiated
2	Initial or experimental implementation (e.g., pilot projects, limited datasets)
3	Operational AI system performing basic tasks with moderate functionality
4	Advanced AI applications in place, demonstrating significant effectiveness
5	Fully integrated, scalable AI infrastructure actively supporting institutional objectives

. The idea here is not to implement a relative assessment (i.e., how do African museums rate versus non-African ones) but rather an absolute one (given the possibilities that AI opens, to what extent African museums make use of them to decolonize curatorial practices). For each axis, repeatable criteria will be used as outlined in all the AI-generated figures, as here in Figure 3.

Currently, the list of AI-enabled museums in Africa includes five institutions: the Digital Dome and Planetarium of Iziko Museums in Cape Town (South Africa, inaugurated in 2017, with AI applications later integrated); Ethiopia’s Science Museum (2022), financed by China but now independently run by Ethiopia; the Dar Gnawa Museum in Marrakech (Morocco, 2024); the Grand Egyptian Museum (GEM) in Cairo (Egypt, 2024); and the International Museum of the Prophet’s Biography and Islamic Civilization (IMPBIC) in Dakar (Senegal, 2025). The reasons for selecting the Dar Gnawa Museum and the Grand Egyptian Museum (GEM) as case studies are as follows. As we already conducted a research on Ethiopia’s Science Museum [38], we decided not to replicate our findings in this present study. The Iziko Museums were also discarded because AI applications were partially grafted to the Digital Dome and Planetarium rather than fully integrated into the eleven Iziko public museums. As for the International Museum of the Prophet, the fact that it was just inaugurated at the time of the writing would compromise the longitudinal depth of our assessment.

The Dar Gnawa Museum was chosen as case study because our Korea-based research team just signed a Memorandum of Understanding (MoU) with this institution under the global multidisciplinary initiative “Decolonizing a Southern Donor: AI Curation and Archiving in Museum” (NRF-2025S1A5C3A02004512), funded by the National Research Foundation of Korea. Through this collaboration, we wish to explore ways of merging cultural heritage with technological innovation. As for the GEM, it stands as a flagship project of more than two decades in the making, setting a benchmark for how African institutions are adopting cutting-edge AI technologies into museum practice. Egypt and Morocco both position themselves as cultural and technological hubs bridging Africa and the Global North. Moreover, the Dar Gnawa Museum and the GEM were selected because, having suffered from colonization and heritage looting—with objects dispersed across numerous museums in the Global North [16,17]—, Morocco and Egypt’s museumscape still reflect “a local history of colonial museology and, at the same time, (…) political ambivalence over the role the museum should play in the nation’s postcolonial life” [19]. As both Morocco and Egypt consider themselves as regional AI leaders with ambitious museum policies [6,7,46], we wondered what role did the above-identified AI functions play in decolonizing curatorial practices and in building synergies between the recovery of cultural and technological sovereignties. To this end, we conducted two rounds of semi-structured interviews with local museum staff and AI professionals, some of which can be found in the Supplementary Material [6,7,46,81,82,104,105]. The first round concentrated on the specific AI functionalities within museums and was guided by the following set of open-ended questions.

• Has your institution fully archived and catalogued its collections using AI?
• Could you describe how AI engineers and curators collaborate within your institution?
• Have machine learning, neural networks, or computer vision been used to analyze ‘deep features’ for research or restoration purposes? Has AI been employed to study or predict visitor behavior?
• In what ways do visitors benefit from AI applications in your institution?
• Does your institution promote artistic creation through AI, whether by visitors or by artists?
• What upcoming AI-related projects are planned for your institution?

As stressed earlier, while necessary to understand how these museums integrate AI, these questions did not sufficiently unveil the potential colonial frictions between these institutions and their partners. Therefore, a second round of semi-structured interviews was dedicated to a broader discussion on the bridges between museums and their AI ecosystem. It consisted of the open-ended questions listed below.

• How has the transfer of equipment and expertise in digital and AI technologies taken place? Were AI-related tools and capacities imported from abroad (and if so, from which countries), or were they developed locally?
• Where is your institution’s heritage data stored and managed? Is it hosted abroad or within national data centers?
• Given that a significant portion of your country’s heritage is located abroad, how do you see AI in relation to the restitution debate? Can it serve as a substitute for material restitutions, or rather as a catalyst for them? In what ways?
• Would you consider AI to be inclusive or exclusive for marginalized groups?
• In your view, does AI contribute to the decolonization of museums, or does it risk reinforcing colonial dynamics? How so?
• How are your institution’s AI initiatives aligned with your country’s national AI policies?
• What solutions is your institution considering to ensure the financial sustainability of its AI ecosystem?
• Your country has been proactive in pursuing AI diplomacy. Are there Pan-African or bilateral projects in the field of AI museum diplomacy?

On the basis of the answers, as well as publicly available museum documents, websites, exhibition descriptions, academic articles and field reports, we conducted qualitative content analysis and comparative cross-case comparison based on the abovementioned AI functionality grid (Figure 3). There are of course some inherent limitations to our approach, the first one being that the Dar Gnawa Museum and the GEM have been inaugurated for just over a year at the time of our writing, reducing the scope for longitudinal observation. Besides, our non-African positionality may lead to involuntary but significant interpretation biases, as we can only offer an external standpoint. Having said that, the lack of visibility of AI African museums in the literature and their emergence as new techno-cultural agencies on the continent encouraged us to proceed with our investigation, whose results are detailed in the next section.

4. Results

4.1. Collections Management

Through AI, the Dar Gnawa Museum has moved beyond static catalogues toward dynamic, digitized archives that reflect the living nature of the Gnawa culture. While the process is still ongoing, it has built an internal system that archives and interprets a wide range of resources, especially those that do not fit neatly into conventional museum categories—Lila ceremonies with songs and dances, trance, ancestral invocation, and healing, guided by a maalem (master musician) and a moqadma (clairvoyant) [6]. These domains, which fall under the scope of Intangible Cultural Heritage (ICH), were traditionally marginalized by colonial museums and ethnographic practices that privileged material culture. At the heart of the Dar Gnawa Museum platform is an advanced multilingual Natural Language Processing (NLP) system, specifically trained on an extensive collection of historical Gnawa rare manuscripts, transcriptions of Lila ceremonies and other oral narratives, traditional poetry and ethnographic records spanning centuries. This proprietary technology can fluently treat Arabic, English, Spanish, and French texts, reflecting Morocco’s rich linguistic diversity while safeguarding the spiritual vocabulary distinctive to Gnawa rituals—ensuring both academic precision and cultural fidelity [106]. By integrating multi-lingual metadata, the Dar Gnawa Museum’s NLP model counterbalances monolingual curatorial frameworks historically imposed by colonialism. During the current phase of the project (2024–2026), the Dar Gnawa Museum is also developing a collaborative platform in collaboration with the Moroccan Ministry of Culture to showcase rare recordings of legends like Hajja Hamdaouia through digital archives. Here, machine learning and AI-enhanced computer vision may accelerate the digital classification, description and tagging of artworks from vast collections. In other words, the process of inventorying Gnawa oral traditions, manuscripts and objects, though not fully completed, is well under way [6].

Compared to the Dar Gnawa museum, the GEM has a much vaster collection of over 100,000 artefacts, some of them of 700,000 years. As such, the GEM is poised to become “the largest museum in the world dedicated to the people, history, culture, and beliefs of Ancient Egypt (with items) from the Prehistoric Period to the end of the Roman Empire in Egypt” [107]. Japan’s International Cooperation Agency (JICA) has played a pivotal role in the GEM’s development, providing two ODA loans totaling roughly $800 million in 2008 and 2016 to fund its construction and facilitate expert trainings. Between June 2008 and June 2011, the Artifacts Database Development (ADD) project, established by JICA and led by an Egyptian team, upgraded the existing database with more detailed artifact records while simultaneously building the institutional and human capacity of GEM Conservation Center to ensure smooth operations [108]. In spite of this, the GEM AI-powered digital documentation system remains incomplete as it still lacks full-scale cataloguing and online access to the entire collection. Contrary to the Dar Gnawa Museum, the GEM has not developed its own AI tools and instead may be prolonging colonial patterns by relying on a string of international partners—JICA, Meta, IBM, the American University of Cairo (AUC), the US-funded Centralized Museum Repository Project, and the EU-funded iHeritage initiative [46]. This could be explained by the fact that, as a public institution, the GEM AI initiatives directly depend on the Ministry of Communication and the National AI Council, who, in addition to preferring foreign capacity transfers, privileged health, education and agriculture over heritage in terms of sectorial AI applications [7].

4.2. Research

The Dar Gnawa Museum managed the feat of designing “Kouyou”, a 100% Moroccan AI that collects data and creates algorithmic heritage, with the ultimate aim of harnessing technology for cultural sovereignty [109]. “Rather than outsourcing expertise, local talents—i.e.: engineers and ethnographers who understand both the technical and cultural nuances—were solicited. Encoding sound, particularly ritual chants and polyrhythmic drumming, posed a real challenge requiring more than just signal processing. It demanded cultural sensitivity first and a willingness to rethink what ‘data’ means in a heritage context” [6]. Kouyou’s training data—developed in partnership with Gnawa Maâlems—draws on rare manuscripts, transcriptions of Lila ceremonies, and traditional poetry. More than just a translation tool, Kouyou uses adaptive algorithms to detect regional linguistic nuances, from Saharan Hassaniya accents to urban Darija variations, enabling deeply localized storytelling that is decentered from Northern narratives and connects with Moroccan youth and international users alike [106]. In its current development phase (2024–2026), the Dar Gnawa Museum is undertaking a comprehensive initiative entitled Digitizing Morocco’s Musical Mosaic. This phase comprises several interrelated projects, including the application of machine learning to transcribe the oral rhythms of the traditional Amazigh collective dance Ahwachi into standardized musical notation [6]. In parallel, the museum is developing NLP models to interpret the language embedded in rural folk songs Al-Aita and Andalusian-influenced classical poetry Al-Malhoun. In partnership with the Regional Center for Saharan Arts in Laayoune, this initiative also focuses on AI-driven voice synthesis to clone the voices of Sahrawi griots. [106]. In sum, the Dar Gnawa Museum is conducting extensive research throughout a finely decolonial agenda.

The GEM is delving into the field of AI art to rethink the way research has been traditionally carried out. Through machine learning, neural networks and generative adversarial networks (GANs), AI art “can serve as both a reinterpretation of historical data and a catalyst for new lines of inquiry, prompting scholars to revisit and reassess their interpretations of the past (…) This process involves feeding large datasets into machine learning models, which then produce visual representations based on patterns identified within the data sets” [107]. Using AI art, the GEM thus seeks to process and interpret ancient manuscripts or visual artifacts, uncovering latent themes or motifs that are not readily apparent to human observers. However, beyond these declarations of intent, we found limited evidence that would support the articulation of a comprehensive research agenda within the museum. Instead of leading the initiative, the GEM is incorporated into national research programs such as the Centralized Museum Repository Project, which seeks to standardize documentation and collections management across major institutions, including the GEM, the Egyptian Museum (Tahrir), and others [46]. Rather than museums themselves, most AI initiatives in the cultural field are targeting higher education and, in particular, “innovative, cross-sectoral, and sustainable teaching skills aligned with developments in digital education” [7]. Moreover, although the GEM collaborates with a broad range of international institutions that store and display Egyptian artefacts, it has largely missed the opportunity to develop bilateral provenance research and multilateral AI-driven platforms addressing objects illicitly removed during the colonial era. Contrary to the Dar Gnawa Museum, the forward-looking adoption of AI is accompanied by an ahistorical approach to collections and colonization within the research field.

4.3. Visitors Management

The Dar Gnawa Museum has experimented with models that track visitor flow and engagement patterns, but to a limited extent as its curator considers that predictive analytics can be useful, but also risk flattening the complexity of human experience. As a result, he prefers using these tools to enhance accessibility and storytelling, rather than to optimize metrics alone [6].

At the Grand Egyptian Museum (GEM), the audio guide integrates a Content Management System (CMS) to manage and deliver diverse audio recordings and images, textual descriptions and multimedia features. The CMS is equipped with built-in analytics and can therefore track visitor interactions to identify content preferences, enabling staff to tailor future offerings for greater impact. Curators can also update the audio guide with new information, correct inaccuracies, or add seasonal material without disrupting the visitor experience [107]. Besides, the Ministry of Communication and Information Technology (MCIT) has launched AI-powered applications for museum visitor engagement, such as a “knowledge mining application” for the Postal Museum, featuring machine learning-driven information retrieval and a virtual assistant to answer visitor queries. These initiatives show a precedent for using AI in enhancing museum access and experience—tools that could feasibly be expanded to institutions like the GEM in the future [46]. However, the GEM Content Management System is short of taking a clear decolonial shift in the sense that it does not allow visitors to identify themselves according to fluid, multi-layered typologies (languages spoken, diaspora ties rather than mere ethnic coding); nor does it include equity parameters to prioritize underrepresented groups.

4.4. Edutainment

The Dar Gnawa Museum stresses that its AI-powered chatbot Kouyou offers interactive conversational capabilities in Arabic, English, Spanish and French that customize visitors’ experience and enhance digital experiences related to Gnawa heritage. Kouyou will soon offer Augmented Reality tours of Gnawa spiritual sites across Morocco, providing an immersive cultural experience [109]. Kouyou can also be used to assist in music education [6]. However, beyond these features, its primary functions remain largely research-focused rather than geared toward enhancing the visitor experience, and it does not currently offer AI-based edutainment applications other than Kouyou’s role as a generator of algorithmic heritage and interactive disseminator of Moroccan traditional music [6].

Edutainment is probably the area in which the GEM made the most advanced and diversified use of AI technologies. The Children’s Museum is a space designed at the core of the GEM to immerse young visitors in the history of Ancient Egypt through an engaging educational experience. By integrating technology-driven edutainment and gamification, the museum features an interactive journey through time [7]. In a setting that nurtures cognitive and emotional growth, children take ownership of their learning trajectories. The center’s programs, rooted in Ancient Egyptian themes, integrate STEM disciplines—such as coding, digital media, robotics, engineering, business fundamentals, and smart technologies—into engaging educational experiences [107]. But the GEM is also providing immersive experiences to young and grown-up visitors through the Tutankhamun exhibition. Digital projection mapping is used to transform the museum into an immersive space featuring AR and VR experiences centered on the King [46]. Moreover, 3D reconstructions and interactive scans of artefacts revive Ancient Egypt for the visitors. To extend the visitor experience, the GEM also arranges for carefully designed “Instagrammable” moments, encouraging social media sharing and increased engagement [107]. Both the GEM and the Dar Gnawa Museum tend to “glorify” their respective cultures rather than foreground histories of resistance, loss, and transformation in colonial and postcolonial contexts. Yet, while the Dar Gnawa Museum conveys the polyphonic traditions of its multi-ethnic spiritual community through performance-based workshops and ceremonies, the GEM’s gamification initiatives project a more monolithic voice, largely detached from contemporary decolonial debates.

4.5. Artistic Creation

At the Dar Gnawa Museum, advanced digital innovation enables both the preservation and creation of musical heritage. Indeed, “unlike static objects, musical traditions evolve and adapt. It is spiritual, communal, deeply tied to identity, and more representative of Morocco’s plural cultural landscape” [6]. In collaboration with VR studios and Amazigh communities in Ouarzazate and Taroudant, it is producing immersive Ahwach performances that bring ancestral traditions to life. Moreover, the museum uses AI voice synthesis to clone the distinctive nasal vocal techniques of Sahrawi griots, in partnership with Laayoune’s Regional Center for Saharan Arts. Another key initiative involves the creation of interactive sound maps that trace Al-Hassaniya melodies along historic trans-Saharan trade routes, offering users a multisensory journey through nomadic soundscapes. Beyond 2030, Dar Gnawa plans to launch AfriSound, a major digital platform hosting over 100,000 traditional African songs, making it one of the most ambitious projects to share the continent’s musical performances [106]. These creation-oriented projects all foster a sense of Pan-African solidarity that challenges the residual structures and narratives of colonialism.

The GEM currently lacks programs that enable visitors or artists to engage in AI-assisted or data-based art creation.

4.6. Extensions

The Dar Gnawa Museum has turned itself into a hub of different activities that by far exceed its premises—such as cooperation with African diasporas and the Pan-African Music-Tech Alliance, all in close cooperation with scientific and educational communities [104]. The GEM has combined its green initiatives with forward-looking strategies for a sustainable tourism sector including handicraft workers, local manufacturers and farmers [15,110], while simultaneously fostering international museum partnerships and aligning with national STEM policy objectives [7,46]. While it is essential to conceive museums not as ends in themselves but as bridges between society and heritage, the extension wedge of our pie chart remains overly broad, making it difficult to clearly differentiate the diverse commercial, diplomatic, and domestic policy domains in which the museums under study are actively engaged. The coming discussion will therefore strive to connect the curatorial functions of the museums to their broader AI ecosystem. But before turning to the comparative discussion, a mere glance at the radar charts (Figure 4 and Figure 5) reveals that both the Dar Gnawa Museum and the GEM have collection-centric proclivities (collection management and research), despite their different focus on tangible Egyptian antiquities and intangible Moroccan music. This seems to reinforce Thiel’s premise that AI technologies are predestined for indexing and research tasks in museums [85]. Sharper contrasts emerge in the remaining categories. Until its current renovation in August–September 2025 [105], the Dar Gnawa Museum placed much less emphasis on edutainment and visitor management than on artistic initiatives and museum diplomacy. On the other hand, the GEM, with high scores in both visitor management and edutainment, seems to corroborate Egypt’s commitment to a human-centered development of AI [46]. But its role as cultural mediator largely eclipses its non-existent creative initiatives.

5. Discussion

The findings of our investigation thus far indicate that the two African AI-assisted museums under study are—to varying degrees according to each institution—implementing the same functional capacities as their non-African counterparts (Appendix A; Figure 4 and Figure 5). While a technological lag might allow coloniality to resurface in other African institutions willing to develop AI functions through ODA or museum-to-museum cooperation in the future, as far as our two institutions are concerned, AI functionalities are on a par with worldwide standards and do not corroborate a North–South AI divide. But the outcomes revealed by our analytical model, though essential for understanding the museum operations, offer no conclusive evidence that the institutions can effectively advance Morocco and Egypt’s techno-cultural sovereignty. In a few instances, the Dar Gnawa Museum integrated AI functionalities with a clear decolonial potential—multi-lingual metadata over monolingual curatorial frameworks; focus on the Gnawa community marginalized in colonial and post-independence contexts; ongoing research programs and performances involving source communities; community support through handicraft sales and sponsorship at festivals; and above all, development of 100% home-made technology [105]. In others, both institutions valued curatorial pragmatism over historical self-introspection. The GEM, in particular, seems oblivious to the history of tomb-raiding in Egyptology [111]. In other cases yet, it may be necessary to move from a purely functional approach to a more process-oriented one so as to unveil the correlations between decolonial curatorship in the strict sense of the term and the colonial threat posed by the broader AI environment. While the Dar Gnawa museums prides itself on having developed a 100% Moroccan AI chatbot, Kouyou, the GEM resorted to different types of capacity transfers for its CMS and AI inventorying tools [107]. Well aware of the risks of data colonialism, Dar Gnawa Museum’s curator, Oussama Elasri, stressed that Kouyou was not a mere innovation, but an act of cultural resistance against global uniformization [112]. On the contrary, the GEM’s staff did not see reliance on foreign partners like ARCE, USAID and Meta, as a symptom of coloniality but rather as a necessary step towards achieving “Egypt’s AI autonomy” [107,110].

Some lingering paradoxes still remain to be fully addressed. The Dar Gnawa museum does not engage with colonialization or coloniality as curatorial themes within its exhibitions, yet fully aware of the threats posed by data colonialism, developed its own sovereign algorithm; as for the GEM, it seems willing to cloud its data through platforms run by the very countries that looted its artefacts, while at the same time claiming to regain cultural and technological sovereignty. It would seem that stigmas of a clouded coloniality transcend the museums’ precincts as they are not automatically correlated to their curatorial functions. Clouded coloniality should hence be unveiled through a broader lens. As Morocco and Egypt embed their respective cultural heritage within AI-driven systems of curation and interpretation, how are they securing its control and ownership through AI tools without exposing what is now becoming an interconnected techno-cultural sovereignty? Specifically, the discussion will seek to answer the following questions. Considering the enduring influence of digital colonialism on infrastructures and regulatory frameworks, what policies have been adopted to prevent cultural estrangement and foreign interference? At a domestic level, are Morocco and Egypt reproducing center-periphery hierarchies, as posited by data colonialism theories, or promoting inclusivity across diverse social groups? In this vein, to what extent do education projects guarantee equitable access to AI-powered tools and meaningful participation in digital transformation? Are public investments and state-led incentives sufficient to reduce reliance on foreign technology providers and business partners? Finally, in light of the inherently interconnected nature of AI networks and infrastructures, are Egypt and Morocco developing a coherent and competitive AI diplomacy to navigate this era-defining technological landscape?

To evaluate how these challenges affect the two institutions under review, the following section will introduce six refined criteria. These criteria do not evaluate each country’s AI readiness, as the Oxford Insights [113], the Egypt National Council for AI [114], and UNESCO [115] do. Instead, they aim to assess the synergies between each museum and its AI ecosystem. In particular, they evaluate the degree of techno-cultural empowerment, in the face of clouded coloniality, the Dar Gnawa Museum and the GEM reached through synergies with the domains of regulations and infrastructures, inclusivity, education, innovation, economic integration, and diplomacy (Figure 6).

5.1. Regulations and Infrastructures

In its continental AI Strategy, the African Union warns against “an AI divide between and within countries and globally” due to, among other factors, data “sourced from developed countries and from non-diverse and non-inclusive developers’ teams” [116]. Immediately afterwards, it encourages member states to establish “an appropriate AI governance system and regulations at regional and national levels”. As of today, 39 out of 55 African countries have enacted data protection laws, with various enforcement levels [117].

Despite the Digital Morocco 2030 blueprint and the expected creation of a National Artificial Intelligence Agency (both approved in 2024), experts signal that Morocco does not have a unified AI strategy [118]. As for Egypt, it already designed two national AI strategies (2021–2025; 2025–2030). As of mid-2025, neither Morocco nor Egypt have adopted a dedicated AI law, though both are armed with a battery of domestics acts on electronic transactions, data protection and privacy, and fintech laws [114,115]. Egypt’s 2020 Personal Data Protection Law (PDPL, no. 151) restricts cross-border transfers and encourages the retention of sensitive data within national infrastructures, while intellectual property rights (IPR) are being progressively reinforced through updated legislation [46]. Likewise, Morocco’s cybersecurity law (05.20) mandates that all sensitive data be hosted within Morocco’s borders, driving significant data repatriation since its application in 2021 [115,116,117]. This clear pivoting toward internalizing critical data is fomented by Morocco’s most powerful supercomputer in Africa and its 26 national data centers, compared to Egypt’s 14. Out of the whole African continent’s data center capacity—(400 MW) by the end of 2025—Morocco accounts for 140 MW, that is 35% of that capacity [118].

But in either country, it is unclear what percentage of data has been repatriated and even local data centers are run by a mix of domestic and foreign operators. Following the enactment of Morocco’s Cybersecurity Law 05.20, priority has been given to securing strategic sectors like government and bank-related data. In this regard, Digital Morocco 2030 states that cloud services managed by Moroccan operators will be reserved for the public sector and Organizations of Vital Importance (OVI) [119]. Likewise in Egypt, the government cloud computing data center inaugurated by President Abdel Fattah El-Sisi in April 2024 serves as a central repository for all ministry data [120]. However, heritage-related data has not been identified as “sensitive” by Moroccan or Egyptian officials [7], leaving it exposed to potential offshore processing, external control, or even commodification.

As a private institution, the Dar Gnawa Museum took it upon itself that all data are stored locally, within certified Moroccan data centers to ensure that cultural data remains under national stewardship and safeguard Morocco’s digital sovereignty. Its curator Oussama Elasri further stresses that “we are aware of the risks of data colonialism and have taken steps to avoid them” [6]. But nothing guarantees that Morocco’s fourteen public museums follow suit. As for Egypt, through the Centralized Museum Repository Project, which includes the GEM, it seeks to gradually enhance oversight of metadata and collections tracking. This central information system strengthens institutional capacity to manage digital representations of its heritage in-house, supporting local control and reducing risks of narrative displacement [46].

Current Moroccan and Egyptian legislations were not crafted to address the recent transformations of AI-era heritage undertaken by the Dar Gnawa Museum and the GEM, namely, the ownership of AI-generated outputs (e.g., voice clones, 3D reconstructions, translated manuscripts); consent and agency of source communities; and long-term archival access. There are therefore concerns that, as Morocco and Egypt graft European legal standards such as the General Data Protection Regulation (GDPR) on their own laws [121,122], they may reproduce the import of European laws upon and after independence. In the heritage field, former colonizers made sure the 1970 UNESCO Convention on the Means of Prohibiting and Preventing the Illicit Import, Export and Transfer of Ownership of Cultural Property remained non-retroactive, blocking restitution claims. And the 2007 United Nations Declaration on the Rights of Indigenous Peoples (UNDRIP) Convention has no law-enforcing mechanisms. Until Morocco and Egypt promulgate AI laws that ensure the domestic storage of heritage data for their national museums, there will be risks that this legislative vacuum prolongs asymmetries of ownership and sovereignty in the heritage sphere.

5.2. Inclusivity

In order to give themselves a postcolonial veneer, many Northern museums have offered multi-vocal narratives on their extra-territorial collections [123,124]. Yet, as Fuchsgruber points out [125], “museum collections do not automatically become more diverse by being described in more diverse ways”. And technologies add a new layer of ambiguity in the ethical becoming of collections. Mbembe argues that Afro-Computation has the potential of transcending Northern dichotomies between objects, values and virtuality [126]. He argues that “the plasticity of digital forms speaks powerfully to the plasticity of African precolonial cultures and to ancient ways of working with representation and mediation, of folding reality (…) Things and objects, the animal and organic worlds were also repositories of energy, vitality and virtuality” belonging to the world of interfaces. Yet, how can the complexity of African identities—theorized in decentralized paradigms based on multi-allegiance networks, “mosaic of peoples and organizations”, “ethnolinguistic fragmentation”, and “intergenerational solidarities” [127]—be channeled and reflected through ethical technologies? And from an AI perspective, how can data reverse the patterns of underrepresentation and domination that have infused the museum culture in Africa? Decolonizing museums through appropriate AI, understood as an extension of appropriate technology, could be mutually reinforcing insofar as both share a moral injunction: challenging epistemic injustice through “a new relational ethics” [73,128]. To this end, the concept of “data solidarity”—i.e., “the willingness to share datasets and resources with others while acknowledging the invisible processes that take place during the creation, production and sharing of data sets” [125,129]—opens the possibility for inclusive AI within museums. In light of these considerations, what stance do Morocco and Egypt adopt in this debate?

Egypt and Morocco have both insisted on democratizing access to culture through AI; and both seek to positions themselves as leader in ethical innovation in Africa [46,106]. Since 2019, Morocco launched an online portal listing 7000 cultural heritage items to promote and broaden access to its national heritage. More recently, the “culture.ma” platform was introduced to further democratize access to cultural content [106]. Meanwhile, Egypt’s AI strategy is underpinned by the Egyptian Charter for Responsible AI. Its minister of Culture “called for international cooperation to establish ethical standards for the use of artificial intelligence (AI) in culture and arts, ensuring content authenticity and respect for cultural contexts” [130]. Translating intention into action, Egypt’s Ministry of Culture established the National Archive for Intangible Cultural Heritage, a digital platform that enables communities to contribute photographs, videos, and narratives while safeguarding intellectual property rights and ensuring public accessibility [46]. In parallel, another research initiative collaborated with Bedouin groups in North-Central Egypt to develop a mobile application for the self-documentation of their intangible heritage, including oral epics and traditional games. By privileging local agency, the model fosters culturally authentic documentation and broadens access for underrepresented rural populations [46]. But Egypt’s wish to make content creation more accessible to marginalized groups—especially women, youth in remote areas, and cultural minorities—has remained aspirational within the GEM where Egyptian culture is presented as a monolithic Arab-Egyptian bloc straddling over the Nubian, Bedouin, Beja, Amazigh, Copt and Dom minorities. In fact, Egypt has been keener on leveraging AI in the tourism industry at large than on designing tailor-made practices for the GEM and museums in general [110,131]. On the contrary, Dar Gnawa Museum [106] is taking concrete steps to “democratize access to Lila rituals, mystical poems, and the history of the Gnawa”. Indeed, after the Royal Institute of Amazigh Culture (IRCAM) was instrumental in granting official language status to Tamazight-the language spoken by the Gnawa community-in the 2011 Moroccan constitution, the Dar Gnawa Museum is now working in partnership with IRCAM and the Ministry of Culture to expand the application of the Kouyou algorithms beyond the spiritual and artistic Gnawa community onto the Ahwach (Atlas Mountains), Al-Malhoun (imperial cities), and Al-Hassania (Sahara) repertoires. By deciphering Morocco’s linguistic landscape, Kouyou’s AI engine mirrors the diversity of its communities. Here, the ethical dimension of Dar Gnawa Museum’s AI practice preempts legal initiatives and lays the foundations of an indigenous peoples right to possess, govern, access, and maintain control over data that originate from them and relate to their communities, cultural practices, knowledge systems, or territories [132].

5.3. Education

As early as 2019, Egypt and Morocco were among the 50 governments that endorsed the Beijing Consensus on AI and Education, adopted by the UNESCO [133], with a special focus on “multi-stakeholder, and inter-sectoral approach”. As for the African Union 2063 Agenda, its second goal calls for “well-educated citizens and skills revolution underpinned by Science Technology and Innovation” [134].

Morocco, in collaboration with the United Nations Development Program (UNDP), already established an International AI Centre in Rabat that host fellowships, hackathons, pilot projects and provides technical support to governments in the MENA region [128]. It has unveiled plans to launch a dedicated AI university by 2026, positioning itself as a continental hub for AI training, research, and advanced education. The institution will offer Master’s and PhD programs in fields such as machine learning, computer vision, AI law, and the application of AI for development [131]. Besides, it is worth noting that 72% of the country’s scholarship recipients are in the fields of Exact and Natural Sciences, particularly in Engineering and Computer Science; and 27.23% of graduates in science, technology, engineering, and mathematics (STEM) [115]. The objective is to triple the number of graduates trained annually in the digital sector, increasing from 8000 to 22,000 by 2027, with an ambitious target of 50,000 graduates by 2030 [115].

As for Egypt, it has stressed that education policies must prioritize youth empowerment [130], while deploring the lack of AI proficiency and basic digital skills among professionals [110]. One of the main strategic orientations of its National AI policy Plans (2019; 2025) is the importance granted to education and universities [110,135]. The Ministry of Communications and the National Council on AI aim at developing innovative, cross-sectoral, and sustainable teaching skills compatible with sectorial applications [7]. To this end, Egypt has deployed four initiatives, namely Senior AI Expert Cultivation and Attraction, International Academic Alliances, Cross Discipline Course Development, and Comprehensive Qualification Certification System [114]. The state-funded Digital Egypt Builders Initiative (DEBI), the Egypt FWD program with Udacity to train 100,000 people in 18 months, and a nationwide network of Applied Technology Schools launched in 2023 are but a few initiatives which show that Egypt is placing AI and digital capabilities at the heart of its education strategies. In this vein, the Egypt University of Informatics (2021), the first ICT-focused university in MENA, is offering advanced education in AI, computer science, and digital design. But how do these macro-level policies translate into museum-level strategies?

The Dar Gnawa Museum has deliberately expanded the scope of its Kouyou software, positioning it as more than a museum tool and turning it into a driver of educational engagement. In practice, Kouyou now serves both “as a digital resource for schools, and as a pilot application for the 2030 Football World Cup aligning with Morocco’s national digital education strategy to make cultural learning more engaging and widely accessible”, offering access to an open-source library of traditional sounds from across the continent [109]. Building on this trajectory, the museum is also backing AI-driven workshops at international cultural gatherings such as the Essaouira Gnaoua Festival and Zanzibar’s Sauti za Busara, while preparing to launch AI Academies in Rabat and Nairobi designed to train African musicians and engineers in cutting-edge heritage preservation technologies through reverse tutelage [106].

What stands out from Egypt education initiatives in museums is a high level of cooperation between domestic and foreign partners. The Ministry of Communications and Information Technology (MCIT), in collaboration with Dell and Egyptian universities, launched the AI Capacity Building Initiative to train professors in AI fields such as data science and engineering, develop curricula, and provide students with workshops, hackathons, and access to high-performance computing through the Applied Innovation Center. This program equips trainees with AI skills transferable to cultural heritage [46]. Through the Digital Heritage Portal, MCIT and the American University in Cairo (AUC) students and staff digitize collections and enhances them with metadata since 2023. At Alexandria University, the Edu-MUST (Education and Capacity Building in Museum Studies) Program offers postgraduate diplomas and master’s degrees in museum studies while developing a VR interface with 3D models of collections. Conducted in partnership with Egyptian universities (Ain Shams, Helwan, Damanhour, Université Française d’Égypte, and the Ministry of Antiquities) and European institutions (University of Southampton, École du Louvre, among others), it directly trains future museum professionals in digital curation, VR interaction, and interdisciplinary heritage technologies [46]. The GEM has been keen on targeting the youngest, pre-university level generations through its Children’s Museum. Through robotics, programming, simulation techniques and serious games it wishes to enable young visitors “to develop invaluable inter-personal skills like problem-solving, logical reasoning, critical thinking and creativity” [107].

Nevertheless, contrary to the Dar Gnawa Museum who developed its own software and extends its use not only in Moroccan schools and museums but also across Africa, the GEM does not break away from a pattern of importing skills and technologies, as evidenced by the national AI workforce development program launched by the Ministry of Communication in partnership with Microsoft and its Azure-based model development. Morocco seems more inclined to rely on its own capacities. By way of illustration, Mohammed VI Polytechnic University established a Digital Interactive Center as a public–private partnership, serving as an innovative academy for training and deploying digital economy professions, particularly in Virtual and Augmented Reality (VR/AR) technologies; but it is not cooperating with museums as of now. With the advent of AI, museums are likely to rely more heavily on universities, but also incubators and private firms—a trend explored in the following two sections.

5.4. Innovation

At the junction between education and the private sector, innovation is vital to the AI ecosystem. In 2023 Morocco allocated 0.85% of its GDP to Research and Development [134]. Though the country does not provide specific figures for public spending on AI, significant initiatives are noteworthy. The International Center for Artificial Intelligence in Morocco is the first UNESCO Category 2 center dedicated to AI in Africa. Established in 2020 within the Mohammed VI Polytechnic University (UM6P), the Center aims to promote applied research, training, skills development, and capacity building in AI. Besides, with two branches in Berkane (2023) and Taroudant (2024), the National School of Artificial Intelligence and Digital Technologies is the first public institution entirely dedicated to AI and digital technologies [136]. The Al Khawarizmi program with a 4,600,000 EUR budget, is structured around sector-specific use cases in big data, health, media, digital economy, but not culture [137].

However, as aptly stressed by the UNESCO [115], there is a noticeable lack of initiatives on digital anthropology and culture, philosophy of technology, or AI ethics. In this respect, the Dar Gnawa Museum is actively seeking partners who understand the cultural and digital stakes and are willing to invest in experimentation, not just outcomes [105]. Irrespective of the lack of state-sponsored AI programs in the cultural field, the Dar Gnawa Museum could benefit from individual researchers’ initiatives. The JEEM Benchmark and Visual-Language Models program, for instance, evaluates vision-language AI models across four Arabic dialects with tasks like image captioning and VQA on culturally rich visual material [138]. Likewise, the DaToBS OCR for Amazigh (Berber) Signs, a deep learning-based system developed to automatically recognize and transcribe Tifinagh script from natural images, facilitates cultural preservation and literacy in the Amazigh language [139]. Such synergies could upgrade the Kouyou algorithm, whose strength is its potential transferability to other traditions (Ahwach, Al Malhoun) [140].

In 2023, Egypt allocated 1% of its GDP to research and development expenditure [141]. Egypt’s AI National Strategy focuses on fostering innovation by promoting the development of indigenous language foundation model to gain autonomy in the field of AI [114]. The Applied Innovation Center (AIC), as the primary implementation arm of the Ministry of Communications and Information Technology’s AI strategy, is actively developing projects across multiple sectors, including culture. Its initiatives underscore Egypt’s commitment to leveraging AI for heritage preservation, digitization, and enhanced public engagement. A research team developed EgyBERT, a large language model pretrained on over 10.4 GB of Egyptian dialect text, primarily sourced from tweets and online forums—the largest corpus of its kind to date. In parallel, the Egyptian University of Informatics created a text-to-speech model in Egyptian Arabic. Designed for applications such as interactive museum guides and accessibility tools, this model also supports the production of audio educational content for visually impaired users. Public GitHub projects such as “ASR_for_Egyptian_dialect” illustrate ongoing efforts to develop Automatic Speech Recognition (ASR) systems tailored to Egyptian Arabic. Leveraging synthetic data, pretrained architectures, and multi-stage training, these initiatives advance the state of speech-based AI and provide a foundation for future applications [46].

In this context, Egypt is introducing a patent licensing system specifically for AI to promote innovation and safeguard intellectual property rights, with the aim of spurring innovation but also drawing in both foreign and domestic investment [114]. Furthermore, Egypt seeks to boost innovation by strengthening “in-depth cooperation between Egypt’s industry, universities and research institutes, from platform start-ups, to research labs and technological transfers mechanisms, aligning research endeavors with industrial needs” [114]. Egypt also inaugurated innovation centers and industry hubs such as the abovementioned Applied Innovation Center (AIC) which fosters pilot projects in AI, big data, HPC, IoT, and cybersecurity to support scalable solutions. As with Morocco, it is worth noting that these innovation centers are not primarily focused on culture, thought their findings may be transferable. Besides, in Egypt’s case, they heavily rely on foreign partnerships (IBM’s EGNC and SkillsBuild programs; Capgemini AI Center for Excellence in Cairo; or the partnership between the government and Microsoft to train and certify 100,000 individuals).

Though not directly related to the GEM, three initiatives bode well for the heritage sector in Egypt. Egypt’s Center for Documentation of Cultural and Natural Heritage [142], housed within the Bibliotheca Alexandrina, uses AI-supported digital documentation, computer vision, and heritage cataloguing technologies to preserve and share Egypt’s tangible and intangible heritage. Moreover, the Creativa Innovation Hubs organized the first Metaverse Hackathon across its multiple university-based hub (in cities like Mansoura, Monufia, Minia, Sohag, Qena, and Aswan), mobilized students, developers, and creators to design digital experiences tied to Egyptian culture and heritage [142]. Last but not least, the MCIT and the Ministry of Culture have spearheaded ambitious digitization initiatives across major cultural institutions, most prominently through Egypt’s Digital Heritage Platform, which integrates collections from the National Library and Archives, the Cairo Opera House, and others [46].

5.5. Business

In 2018, Google, IBM and Meta all inaugurated AI labs in Ghana, Kenya and South Africa, and Nigeria, respectively. In 2023, Google for Startups Accelerator granted ten startups from South Africa, Ghana, Senegal, Uganda, Ethiopia, Kenya, and Nigeria a total of $8.488 million USD in funding [53]. Mazzanti et al. note that, for museums, the benefits of investments in AI “often remain inaccessible to smaller institutions due to the high costs, limited funding, and the perception of AI as being non-essential” [4]. In the absence of frugal innovation from local businesses, most African museums therefore remain in a state of dependence [52]. Here, the ability for museums to adjust to the AI ecosystem depends to a great extent to their ability to create partnerships with the private sector [115]. Following its Digital Morocco 2030 strategy [119], the Kingdom is creating incentives for Income Tax (IR) and Corporate Tax (IS) through the start-up label as well as other incentives like life grants, incubation grants, honor loans and seed loans and subsidies for digital transformation projects. But once again, as noted by the UNESCO [115], the sectors that are receiving significant investment in Morocco are not culture but healthcare and biotechnology, media, marketing and social platforms, as well as IT infrastructure. Here, Morocco must tread a thin line between protecting its own interests and companies, and attracting foreign capitals and skills to upgrade its AI sector, at the risk of undermining its techno-cultural sovereignty. While the education ministry (MESRI) and the digital transition ministry (MTNRA) have already established partnerships with Oracle and Orange [115], Morocco is in parallel seeking to retain the initiative on the African continent. In 2023, 2024, and 2025, for instance, it held the GITEX Africa, a major technology and innovation event that brings together government officials, tech giants, startups, investors, and researchers to showcase and discuss the latest developments in AI, positioning Morocco as a regional hub for exporting AI compute services in and beyond Africa.

As an independent institution relying on sponsors and private partnerships, the Dar Gnawa Museum is experiencing financial difficulties. So far, the museum has been sustained almost entirely through limited private and personal resources, driven more by ambition than by institutional support. There is no stable funding model in place yet, which makes long-term planning precarious [105]. By partnering with Casablanca fintech startups to tokenize traditional music and ensure royalties return to communities (e.g., Tiznit’s Ahwach troupes), the latest museum’s initiative acts as a catalyst for inclusive growth—creating skilled jobs in AI, heritage tourism, and creative industries while supporting the New Development Model and Generation Green 2030′s vision of a resilient, identity-driven, and innovative economy. But as a note of caution, the curator stresses that “the ultimate goal is to support creativity, not to commodify it” [6].

Under the Egyptian Ministry of Communications and Information Technology (MCIT), Egypt’s national AI strategy aims at raising ICT’s contribution to GDP to 7.7% by 2030. Yet, integrating AI in businesses, comes with a number of challenges, especially in terms of the unaffordability of digital transition for small and medium-sized enterprises (SMEs) [110]. To achieve this goal, Egypt’s AI strategy (2025–2030) has created an AI venture fund to give financial support to start-ups, a fund to which governments, the private sector, and overseas investors can all contribute. In addition, Egypt has established AI incubators and accelerators to offer specialized services to startups-from office space and technical support to entrepreneurial advice and market docking [114]. In parallel, Egypt is currently working with international partners such as NVIDIA, the Gates Foundation, and UNDP [131]. To celebrate the 122nd anniversary of the Egyptian Museum, the “Revival” initiative was launched in November 2024. Born out of a collaboration between the Ministry of Tourism and Antiquities and Meta, this digital innovation initiative resorts to Instagram’s augmented reality (AR) filters to virtually restore and animate artifacts from both the Egyptian Museum and the National Museum of Egyptian Civilization (NMEC). This initiative aims to deepen public engagement and advance the digital preservation of Egypt’s cultural heritage [110]. Besides, through its architecture and operational model, the GEM incorporates commercial areas featuring Egyptian-owned shops and restaurants, establishing itself as a contemporary cultural and retail hub that blends heritage display with economic activity [142]. The museum is also envisioned as a tool contributing to sustainable tourism by channeling visitors toward the nearby archaeological sites of the Memphite Necropolis. It also plays a key role in supporting local industries and handicrafts, such as papyrus workshops in Giza and the renowned weaving tradition of Al-Harraniya Village, known for its carpets and kilims. The Harraniya modern art school further enriches this landscape with batik wall tapestries and ceramic wall hangings. Yet, these heritage crafts face the threat of decline, as artisans increasingly abandon them in favor of other professions, reflecting both economic pressures and the fragility of traditional knowledge transmission [15]. Having said that, compared to the Dar Gnawa Museum, the GEM enjoys a more stable business model which beyond public funding is also more exposed to and reliant on foreign capital. In facing frictions between clouded coloniality and techno-cultural empowerment, both Egypt and Morocco have turned to AI diplomacy—through their museums too—a trend to which we now turn.

5.6. Diplomacy

Given the interconnected dimension of AI, museum diplomacy is a key instrument to wield influence and strengthen museum networks on the international stage. In 2019, the originally non-AI-driven Dar Gnawa Museum managed to inscribe the Gnawa music on UNESCO’s (2019) Representative List of the Intangible Cultural Heritage of Humanity. This international visibility enabled the Dar Gnawa Museum to attract the attention of African diasporas. Since its AI reconversion in 2024, the Dar Gnawa Museum is part of a broader ecosystem that is being currently structured under the banner of SMIALLIANCE, a certified initiative based in Canada, dedicated to bringing together stakeholders of intangible heritage in Africa and the diaspora around innovative, inclusive, and sustainable tools. This ecosystem notably includes: SMIAart, a label dedicated to the recognition and transmission of heritage know-how with high cultural and creative value; SmiAfrica, a pan-African network for cultural and technological cooperation, based on the pooling of resources, local expertise, and community dynamics; and Kouyou, the intelligent museum assistant, designed as an ethical tool to support mediation, documentation, and participatory promotion of living heritage [104]. The museum has successfully adapted the different facets of its diplomacy to each geographic area in Africa. In the Sahara region, it built a partnership with Laayoune’s Regional Center for Saharan Arts to clone the voice of Sahrawi griots; and it is collecting and preserving nomadic soundscapes linking Al-Hassania melodies to historical trans-Saharan trade routes. In the mid-term (2026–2030), it aims at building a Pan-African Music-Tech Alliance in West Africa, East and Southern Africa based on reverse tutelage. Together with the National Museum of Bamako and Dakar’s Ecole des Arts, it is adjusting Kouyou to preserve, respectively, Mali’s traditional griots (Mandé jeli) and Wolof mbalax rhythms. Supported by Mombasa’s Swahili Cultural Center and the AU’s African Music Council, it will map the Arab-African fusion of Taarabl music and compare Moroccan Sufi chants with South African gospel harmonies. Its diplomacy through music initiative also envisions to turn Morocco into the headquarters for a UNESCO-backed AI for Intangible Heritage initiative and to connect African and global artists for AI-assisted fusion projects [112]. The Pan-African dimension of the Dar Gnawa Museum diplomacy is therefore in line with the African Union Agenda 2063′s Aspiration 5, which envisions “Africa with a strong cultural identity, common heritage, shared values and ethics” [143].

The GEM is far from lagging behind when it comes to museum diplomacy. Working alongside the National Research Institute for Cultural Properties in Tokyo, JICA spearheaded extensive capacity-building programs between July 2011 and March 2016 to advance conservation expertise and strengthen the GEM Conservation Center’s position as a leading regional hub for conservation research and practice. More than 100 specialized training sessions were delivered, benefiting some 2250 participants across various conservation disciplines [108]. In February 2024, the Grand Egyptian Museum earned the distinction of becoming the first museum in Africa and the Middle East to obtain the EDGE Advanced Green Building Certification from the World Bank’s IFC, with support from SECO (Switzerland) and the UK Government. Incorporating measures such as reflective roofing, high-efficiency lighting, intelligent shading systems, and water-saving installations, the museum has cut its energy consumption by more than 60% and reduced water use by 34% [144]. The GEM has also engaged in AI-focused cultural diplomacy: its Learning Center director, Gehane Nabil, participated in the UNESCO High-Level Forum for Museums in Hangzhou (April 2025), which brought together over 190 cultural leaders from 60 countries to discuss AI’s potential in collection management, conservation, and personalized exhibitions, as well as ethical concerns such as copyright, bias, and human oversight [46]. It has also developed its own strand of museum-to-museum cooperation with Paris Centre Georges Pompidou to enhance AI-power audio guide services; and with the ARCE (American Research Center in Egypt) Database Museums Project to modernize the management of its collections [107]. In parallel, Egypt plays an active role in regional AI governance, chairing both the African Union Working Group on AI, tasked with developing a unified continental strategy and joint capacity-building projects, and the League of Arab States (LAS) Working Group on AI, which seeks to establish shared guidelines and regional frameworks across Arab states [46].

Morocco and Egypt’s museum diplomacy illustrate distinct expressions of AI diplomacy. In the case of Morocco, the Dar Gnawa Museum grounds its diplomatic role in the promotion of formerly marginalized identities to foster Pan-African and diasporic alliances. As an independent institution, the Dar Gnawa museum seeks to compensate for limited governmental support with private partnerships, and thus advances a bottom-up model of engagement. This configuration suggests that the museum is resisting clouded coloniality by crafting a form of techno-cultural empowerment model that derives from diverse stakeholders rather than from state-driven initiatives alone. By contrast, the GEM, while enjoying strong governmental backing, has pursued a more North-oriented trajectory by aligning itself with Japan and with a string of Northern-based corporations such as IBM, Meta, Microsoft, Capgemini, among others. In doing so, the GEM appears less attuned to the paradox that its AI-driven heritage is, to a greater extent than the Dar Gnawa Museum, being modelled but not entirely managed by external stakeholders. This dynamic reinforces our argument that a strand of decentralized neo-coloniality insidiously operates under the guise of technological solutionism—what we conceptualized as a friction between clouded coloniality and techno-cultural empowerment in African AI-enabled museums (Figure 7 and Figure 8).

6. Conclusions

This article provides a comparative analysis of the Dar Gnawa Museum and the GEM. It contributes to the literature by introducing two analytical frameworks that assess the AI functions within museums and their degree of techno-cultural empowerment in a given AI ecosystem. As far as curatorial functions are concerned, the findings reveal that, despite their respective focus on intangible and tangible heritage, both museums are collection-centric, thereby confirming Thiel’s hypothesis according to which AI technologies are inherently predisposed to indexing and research tasks [85]. Both institutions pursue extensive collection management and research agendas, less so the GEM, perhaps due to the fact that its official inauguration was in November 2025. At the same time, the GEM stands out as more visitor-centric, operating a Content Management System (CMS) and offering a broader range of edutainment activities, which the Dar Gnawa Museum is only set to expand after its renovation in August–September 2025. Contrary to the GEM, the Dar Gnawa Museum resorts to AI for artistic creation initiatives, a dimension absent at the GEM. Both institutions, however, engage in diverse educative, commercial and diplomatic activities that extend their reach far beyond the museum grounds, underscoring their environment-centric orientation. Overall, the Dar Gnawa Museum and the GEM feature no significant gaps in AI functionalities compared to non-African AI institutions, yet their state-of-the-art technological capabilities do not necessarily protect them from a clouded form of coloniality which transcends conventional curatorial content.

This paper cautions against the longstanding ahistorical bias that has treated technological and cultural sovereignties as distinct domains. At a policy-making level, the import of museum infrastructures and equipment long maintained African museums in a state of dependence; today, given the high costs of AI technologies and the modest budgets usually allocated to culture, museums are likely to become ever more dependent on public–private partnerships. While corporate sponsorship in museums is hardly new, the scale and scope of private-sector involvement may expand to the point of redefining the very concept and practice of technology transfers in museum cooperation. In the broader ODA landscape, policy-makers have already called for a shift away from traditional aid toward “a wider range of financial instruments and South–South cooperation” [145]. This reorientation further stresses the need for apprehending museum cooperation through new theoretical lenses.

At a theoretical level, coloniality, as we have sought to re(de)fine its acceptations, may not be as clearly identifiable as a mere imbalance in technological capacity may suggest. Where museums and AI converge, coloniality may indeed be concealed under a clouded veneer. Ceding AI capacities to external actors effectively entails the deterritorialization of heritage management. In a literal sense, heritage data is clouded and therefore potentially stored, managed, or worse, owned by foreign stakeholders. Indeed, AI is often mistakenly considered as a mere tool where the promise of cultural empowerment acts as a smokescreen behind which technological dependence and hence coloniality might creep in. As the GEM and Dar Gnawa Museum cases show, this strand of coloniality may furthermore be clouded by the fact that it is decorrelated from exhibition content and not necessarily exerted by former colonizers but by a transversality of fields and actors with whom local governments cooperate, albeit at the expense of their very museums and heritage.

Mostly aware of these notes of caution, the Dar Gnawa Museum and the GEM have designed singular strategies to pursue techno-cultural empowerment. The GEM is using AI to empower the youth through its Children’s Museum and SNS-friendly activities. Moreover, Egypt’s ambitious AI policies are gradually bringing R&D and the private sector closer to the heritage field. In this respect, the GEM has been conceived as a launchpad for AI in the tourism industry. As for the Dar Gnawa Museum, not only did it design Kouyou to collect and generate its own algorithmic heritage, but Kouyou’s primary strength lies in its transferability to other communities within and beyond Morocco. To further safeguard the Gnawa heritage, all of the museum’s data is stored in certified national data centers. Most promisingly, the Dar Gnawa Museum is advancing an ambitious AI diplomacy that places strong emphasis on accessibility, reverse tutelage, and creativity on a Pan-African scale. In this sense, unlike the imported museums of the colonial and post-independence eras, the Dar Gnawa Museum and the GEM function less as threatening mirrors of a clouded coloniality shaped by external actors with encroaching interests than as loci of techno-cultural empowerment.