The structure of the new AI world is being designed and built today yet large parts of the global south are not among its architects. Along with the rest of the Global South, Southeast Asia finds itself positioned in the global AI economy as a consumer, a reservoir of natural resources and cheap labor, and the primary supplier of the data powering the Fourth Industrial Revolution.
Beyond recent, progressive data regulation efforts, the crafting of national AI strategies, ministerial declarations, and local “unicorns” built on foreign cloud infrastructure, lies a foundational question: are these digital instruments of technological sovereignty and economic productivity, or rather symptoms of intensifying dependency, democratic erosion, and foreclosure of the region’s techno-scientific future?
Like Dante’s descent into successive circles of hell, this article moves through layers of increasing structural depth, each darker and more unsparing than the last. Guided by history and political economy, I travel downward from the surface layer of policy, compliance frameworks, and AI adoption, through AI’s physical infrastructure, and the deep tectonic plates of the stratified and unequal world-system.
This epistemic movement reveals a historical pattern. From the colonial plantation to the data center, from the steam engine to the GPU cluster, every industrial revolution returns to an unresolved techno-societal puzzle: how does technology shape the relation between society, labor, productivity and surplus, and how does surplus shape the social body and its institutions?
For Southeast Asia’s 700 million population, the dilemma brought about by this AI wave is the latest iteration of the perennial question that has haunted humanity since the first industrial revolution in the 18th century.
AI’s Transformative Potential?
The sociotechnical infrastructure of Southeast Asia is increasingly anchored to a rapidly expanding ensemble of digital and AI technologies. Large language models (LLMs), computer vision systems, facial recognition tools, and embedded predictive recommendation engines, and so on are now deployed at scale. These technologies are increasingly employed and operationalized across e-commerce platforms, public service delivery, financial systems, border control, education, policing, and surveillance networks. Their cumulative effect is a fundamental reorganization of the state apparatus, social relations, urban infrastructure, and the economics of the region.
The dominant interpretation of this transformation is one of techno-progressivism and “charting paths into a brave new world.” Techno-optimists have been quick to announce that Southeast Asia will be “a burgeoning hub for AI innovation” one that has already attracted $30 billion in AI infrastructure and $50 billion investment in AI. The ASEAN ministerial meeting on Science, Technology and Innovation (AMMSTI) released a statement in June 2024 declaring that AI has “significant transformative potential,” and is a “key driver of technological advancement and innovation.” The meeting noted that AI may possibly result “in a 10 to 18% GDP uplift, valued at approximately USD1 trillion by the year 2030.” The market appears to share this worldview, with a report in The Business Times in March 2026 announcing that “81 per cent of South-east Asian companies [are] already piloting and scaling AI-powered projects.”
Yet the picture is more complicated. Scholars and AI experts warn that this wave of “automation is outpacing ethical checks and balances” and will only intensify structural “fault lines,” the “democratic deficit” and economic inequalities. These critics argue that the orthodox narrative reproduces, almost without deviation, the worldview, metrics, categories and development blueprint of Silicon Valley. The United Nations’ Human Development Report (2025) argues that AI risks enforcing a “future ruled by inequality and eroded freedoms,” with Southeast Asia most exposed. The report indicates that the lower-income countries in the region lack the infrastructure, human capital, and technical know-how for any meaningful gains from the technology.
Policy responses have followed but by no means have solved these contradictions. At best reactive and piecemeal, individual states and regional bodies have recently attempted to counter these dynamics through reskilling programs, national AI ethics guidelines, and regional cooperation mechanisms, each falling short of the pace and scale of the transformation they seek to govern. More concretely, Thailand, Singapore, Indonesia, Malaysia, the Philippines, and others have each attempted to assert territorial sovereignty over their digital resources through data localization laws, digital governance frameworks, Personal Data Protection Acts, and national AI strategies, driven by the recognition captured by the U.S. Department of State’s Pax Silica Declaration (2025): “If the 20th century ran on oil and steel, the 21st century runs on compute and the minerals that feed it.”
At the regional level, the ASEAN Framework on Personal Data Protection (2016), the Framework on Digital Data Governance (2018), and the “Guide on AI Governance and Ethics” (2024) “reflect the growing efforts to standardize privacy compliance.” Yet as Hpone Htoo writes, “iterations of the guide remain non-binding and only highlight best practices and recommendations with no enforcement mechanisms. Adoption remains voluntary and does not supersede any national legislation.” The contradictions run deeper than any of these instruments can reach. For Amazon Web Services, Microsoft Azure, Google Cloud, and OVHcloud, it is business as usual and at full speed. The cloud platforms, foundational models, and data architectures restructuring the region’s economy, social life and nature continue entirely undisturbed.
The question is what that digitalization is built on, by whom, and for what purpose.
Technological Sovereignty Simulacrum
Sociologist and media studies experts, Couldry and Mejias (2019), sought to make sense of this datafication of life, social relations and economy by placing it in the wider history of technology, economy, and geopolitics. For them, this “large-scale social, economic and legal transformation” which reifies and compresses concrete reality by datafying it, is part of a larger structural shift they call “data colonialism.” This is not a metaphor but a techno-societal condition of extraction and appropriation that enforces itself via new economic, legal, and regulatory orders. A variegated network of multinational corporations spanning device manufacturers, platform builders, data brokers, and digital infrastructure providers operates as what they call the “social quantification sector.” This capitalist assemblage turns daily actions, interactions, behaviors and institutional operations into data points that are collected, aggregated, repackaged, monetized, and fed back into the circuits of capital accumulation as for-fee services.
This “quantification” reflects a predetermined political economy. Across AI supply chains, mass mineral extraction from peripheries, semiconductor fabrication plants, ever-increasing data mining at scale, data center proliferation, and the undersea fiber optic cables carrying the data across continents, the entire physical apparatus is privately owned by BigTech and weaponized at will by the United States. Nvidia, Amazon, Microsoft, Google, Meta, and OVHcloud monopolize and centralize this techno-societal grid.
The consequences are visible in the Southeast Asian region’s own landscape. Toothless AI and data governance policies, token native executives leading BigTech, unequal bilateral investments, local “unicorns” built on foreign cloud infrastructure, and the proliferation of mostly foreign-owned 631 data centers across Southeast Asia have produced what can only be called a “technological sovereignty simulacrum.” In Baudrillardian sense, it is a simulacrum not because sovereignty was attempted and failed, but because the structural conditions for its existence have been pre-emptively suspended. What remains is a self-referential image: data centers without ownership, AI strategies without productive or governance capacity, digital governance without infrastructure. It does not point to any sovereign reality. It replaces it, and the state administers the image and its consequences only. BigTech is not only selling the data centers, the AI systems, digital platforms, and the techno-utopian vision. They are also selling the experience of technological sovereignty to states for whom the real thing has been made structurally impossible.
The Machine and the World System Problem
Southeast Asia’s technological predicament is far from new or unique. David Ricardo, in “On The Principle of Political Economy and Taxation” (1817), retracts his earlier views on impact of technology on society. Ricardo had previously argued that machinery benefits both capitalist and laborer alike, the former “by making great profits,” the latter by gaining “the means of buying more commodities with the same money wages.” The logic was classically liberal in its optimism: a positive-sum ontology in which the expansion of productive capacity generates a rising tide, lifting all alike. Later, in his chapter “On Machinery,” Ricardo confesses “I am convinced, that the substitution of machinery for human labor, is often very injurious to the interests of the class of laborer.” He traces this epiphany to Adam Smith’s observation that while the worker’s appetite is bounded by the biophysiological limitation of the stomach, the machine owner’s hunger for wealth and property “seems to have no limit or certain boundary.”
Channeling Smith’s pin-factory example where productivity rises but the worker becomes de-skilled and intellectually diminished, Ricardo draws two conclusions: first, that the machine increases the net revenue of owners of the means of production and makes workers redundant; second, that if capital cannot extract maximum net revenue domestically then the machine “will be carried abroad, and this must be a much more serious discouragement to the demand for labor…by exporting it to another country, the demand will be wholly annihilated.”
Charles Babbage, the technologist and “father of the computer,” in his “Economy of Machinery and Manufactures” (1832), demonstrated that machinery does not increase general welfare but decomposes the labor process into its simplest components, concentrating surplus in the hands of those who own the machines. Marx, drawing on Babbage in “Capital” (1867), goes further: machinery under capitalism is not a neutral instrument of production but a weapon destroying the collective power of the workers. He concedes that technology, under different social relations, relieves arduous labor, enables self-actualization, and opens new domains of creative work. Under capitalism, however, what Ricardo treated as a subjective immoral option (the possible export of machines abroad to not pay labor), Marx reveals as structural necessity. Capital does not choose to go abroad; it is compelled to.
Building on this tradition, the Dependency and World System theorists like Amin, Marini and Wallerstein showed that the global state-market system is stratified and organized around a set of iron laws of which monopoly of technology was central. These laws include the unequal exchange between core economies that own and operate machine production and the peripheral circuits that export raw minerals and agricultural commodities; the transfer of surplus value is such that the periphery is structurally unable to own, create, and produce its own techno-scientific infrastructure.
What does this political economic detour teach us about Southeast Asia’s AI and technology dilemma?
First, Southeast Asia’s structural and technological position is not a developmental lag. The iron laws of unequal exchange, peripheral de-industrialization, surplus transfer, and super-exploitation of labor ensure that the region remains a consumer and provider while the techno-scientific systems and its material surplus accumulate in the industrialized world.
Second, the “technological sovereignty simulacrum,” with its AI and digital governance tools, is an ideological obfuscating mechanism that presents the illusion of control over AI and digital platforms that are materially owned by BigTech. The failure is guaranteed; this is the structural outcome of the first point.
Third, Southeast Asia’s question regarding AI and technology is a novel articulation of the same historical question that political economists and technologists posed across every industrial revolution: who owns the machine, who works it, who extracts the surplus, and who bears the cost. From the steam engine to Ford’s assembly line, from the nuclear reactor to the GPU cluster, this techno-societal puzzle remains unresolved.
AI and digital platforms are accelerating and deepening the crisis, and Southeast Asia’s 700 million inhabitants are, today, on its receiving end. Yet historically, specific structures of accumulation and dispossession are neither permanent nor immutable. The unipolar Silicon Valley-centered order is now fracturing. China’s emergence as a full-spectrum techno-scientific power, controlling mineral processing, deploying its own semiconductor architecture, and extending state-directed digital infrastructure across the region, introduces a genuine challenge to Western technological hegemony. China’s own developmental trajectory opens different possibilities for the Global South.
The conditions for alternative human-centered techno-scientific projects must be a subject of urgent scholarly and policymaking inquiry.