AI, industrial sovereignty and Pax Silica

A version of this was first published at Ang Masa. See also statement by Partido Lakas ng Masa (PLM, Party of the Labouring Masses) further below, “Reject Pax Silica and the Philippines’ transformation into a hub of imperialist war and militarised AI infrastructure.” 

The United States, along with 14 other high-tech countries, established Pax Silica in December 2025 as a “strategic initiative” to counter China’s strength in semiconductors, artificial intelligence (AI)1 and high-level technology (HLT). It seeks to do this through US control of supply chains — from critical minerals, energy and logistics to semiconductors, advanced manufacturing, AI infrastructure, software platforms and frontier AI models.

Pax Silica is a US-led reorganisation of global production driven by geopolitical rivalry and the fusion of industrial policy with military strategy. It is an attempt to build “trusted ally” supply chains that limit China’s access to advanced technologies while integrating partner countries into segmented roles within a US-aligned technological bloc. Rather than a single agreement, it operates through infrastructure investments, supply-chain restructuring, security arrangements and industrial partnerships.

The Philippines joined the initiative in April 2026. The Philippines is being incorporated via mineral agreements, semiconductor expansion, AI-linked industrial corridors, logistics projects, nuclear energy development and deeper defence cooperation. Under the initiative, the Philippines and the US will establish a 1619-hectare industrial/AI hub in the Luzon Economic Corridor — an “economic security zone” — to shore up US supply chains.

At the centre of this architecture is the Enhanced Defense Cooperation Agreement (EDCA), which enables rotational US military access, prepositioning of equipment, logistical integration and the development of dual-use infrastructure connected to wider Indo-Pacific strategic planning. In this framework, economic integration is layered onto existing military architecture.

This paper argues that the Philippines risks becoming a low (or at best mid-) level semiconductor processing node, a logistics corridor integrated into US military supply chains, a supplier of critical minerals and technical labour, and a forward-positioned territory within broader US Indo-Pacific security strategy.

The US request to place the hub under US law and grant diplomatic immunity to US personnel is a clear indication of the military underpinnings of this venture. The arrangement also aligns with key Philippine elite interests by attracting investment, strengthening export sectors, and reinforcing political and security ties with the US. The initiative is being hailed as an opportunity for HLT-based development.

The question posed, therefore, is how the rise of HLTs such as AI enables national industrial development grounded in sovereignty, technological self-reliance, and sustainable development in the Philippines. Will HLTs become a tool for national industrial transformation and social development, or will it function as a mechanism for deeper dependency under a global techno-military order?

HLTs under capitalism

HLTs are not neutral. They are deeply embedded within imperialism’s military-industrial complex, global supply chains and class structures of ownership. Within the framework of global capitalism (or imperialism), the global AI industry combines highly advanced computing infrastructure in major capitalist powers with vast reserves of precarious digital labour in low-wage countries, reproducing unequal international divisions of labour facilitating the significant transfer of value to HLT and imperialist countries.

Neither is AI a “new economy” that functions autonomously from capitalism. AI models cannot function and produce a single output without an enormous expenditure of labour — human-created data; software engineers to design highly complex system architectures; human-produced infrastructure, such as electricity, to run server farms; and human maintenance and “hidden” human labour in low-wage countries, such as the Philippines, to clean, label, sort and verify data used for training algorithms (machine learning). AI under capitalism is a part of the “constant capital” or “dead labour” that transfers value, not creates it, following Karl Marx’s labour theory of value.

The central issue is therefore not simply whether the Philippines adopts AI and HLTs, but under whose control these technologies are developed and toward what developmental objective they are directed.

Structural constraints 

The Philippines possesses several partial advantages within the global technology economy. These include participation in electronics assembly, an established IT-Business Process Management and service-sector workforce, substantial mineral resources such as nickel, and emerging semiconductor back-end capacity involving assembly, testing, packaging and electronics manufacturing services.

At the same time, the country remains structurally weak in critical areas necessary for sovereign industrial development. The Philippines lacks a strong heavy industrial base, machine-tool industries, energy sovereignty, advanced semiconductor design and fabrication capabilities, robust public research and development systems, and coordinated long-term industrial planning. This creates a core problem for the Philippine economy: the country participates in global technology chains without controlling its highest-value and most strategic segments.

HLTs and sovereign development requirements

AI as a transformative HLT

HLTs such as AI possess significant transformative potential. When combined with robotics and advanced industrial systems, AI can enable industrial automation, productivity growth, logistics optimisation, infrastructure coordination, semiconductor design, advanced manufacturing, disaster prediction, climate resilience, healthcare modernisation and improvements in agricultural productivity. These technologies are increasingly becoming foundational to industrial development across the world economy.

However, these capabilities are not autonomous; they depend upon deeper material foundations. AI is fundamentally infrastructural, with energy, land and water functioning as first-order constraints on its development and deployment. In the Philippine context, these constraints are particularly acute: electricity prices remain among the highest in Asia, energy security is weak, grid fragmentation persists, and infrastructure vulnerability to climate disasters is severe. At the same time, land reform remains incomplete and, in key instances, subverted, further complicating the territorial basis for large-scale industrial and digital infrastructure.

Energy demands and infrastructure

Because AI workloads and high-performance computing generate enormous heat, cooling has become one of the biggest technical and energy challenges in modern data centres. Large-scale data centres and semiconductor ecosystems require stable baseload electricity, high-capacity transmission systems, advanced cooling infrastructure, water supplies and water infrastructure, and resilient logistics networks. For hyperscale AI infrastructure operated by companies such as Google Cloud, Microsoft, Amazon Azure and Meta, cooling infrastructure can become as large and capital-intensive as the computing infrastructure itself.

At present, the Philippines does not possess the degree of energy sovereignty necessary, as demonstrated by the energy crisis gripping the country, for independent large-scale AI-industrial development. As a result, expansion in this sector risks deepening structural dependence on private energy oligopolies, imported fuels and externally financed grid development.

Pax Silica and the civil-nuclear energy agenda

Energy is a key HLT infrastructure requirement. The focus of Pax Silica is the development of “civil-nuclear energy”. This includes, based on the US-Philippines “123 Agreement” on nuclear cooperation: deployment of US-designed small modular reactors (SMRs) by Meralco; the establishment of a nuclear reactor control room simulator and training hub; and partnerships between Philippine universities and overseas institutions, such as Texas A&M University and King’s College London. This nuclear program aims to integrate the Philippines into US-linked nuclear technology and supply chains. Renewable energy such as solar and wind are not mentioned, even in relation to energy security and therefore sovereignty.

AI as an infrastructural dependency chain

AI systems further require integrated industrial ecosystems: semiconductor design and manufacturing capacity, stable and affordable energy systems, hyperscale data centres, high-capacity telecommunications and fibre networks, scientific and engineering talent pipelines, and secure access to critical minerals and industrial inputs. Without control over these underlying foundations, AI development becomes dependent on outsourced computing systems, foreign cloud platforms and externally governed digital ecosystems dominated by major technology powers such as the US, Japan, the European Union and China.

Within this configuration, AI development is not simply a technological transition but an infrastructural dependency chain. The absence of sovereign control over critical inputs means that digital-industrial expansion is structurally mediated by external capital, fuel supply volatility, and privatised generation capacity. This shapes not only the cost structure of AI deployment but also its strategic autonomy, embedding technological development within broader patterns of dependency.

“Trusted partnerships” and emerging AI blocs

In this light, the discourse of “trusted partnerships” under frameworks such as the World Economic Forum (WEF) is useful to examine. This reflects an emerging language of “AI sovereignty” that emphasises secure supply chains, allied infrastructure, and coordinated compute ecosystems. This closely parallels the strategic logic of Pax Silica: the formation of geopolitical technology blocs organised around trust, security and interoperability.

For the Philippines, however, these partnerships are mediated not only economically but also militarily through arrangements such as EDCA and broader US strategic integration. The key issue is therefore not partnership per se, but hierarchy: the Philippines does not negotiate from a position of technological parity, but from within a structured asymmetry of power.

Under such conditions, the Philippines risks participating in AI development only as a subordinate service provider within foreign-controlled technological systems. As a result, technological modernisation under present conditions risks reproducing dependency rather than overcoming it.

Comparative development models

Taiwan: Industrial upgrading through state-led industrial policy

The experience of Taiwan demonstrates that technological advancement is possible through sustained state-led industrial policy, strong STEM education systems, strategic protection and upgrading of domestic industries, and coordinated technology transfer mechanisms. Through long-term industrial planning, Taiwan successfully moved from low-end assembly operations toward global leadership in semiconductors and advanced electronics. Institutions and firms such as Taiwan Semiconductor Manufacturing Company, Acer and Foxconn emerged from this process of industrial upgrading.

However, Taiwan’s success also remains structurally tied to deep integration within US-led supply chains, dependence on global export markets, and geopolitical exposure within the broader US-China rivalry. Taiwan illustrates both the possibilities of technological upgrading and the vulnerabilities created by dependence on externally structured geopolitical and economic systems.

Cuba: Scientific sovereignty under constraint

Cuba’s experience presents a different developmental model. Following the 1959 revolution, Cuba developed a state-led scientific system centred on universal education, centralised research institutions, public health-oriented innovation and scientific planning. Institutions such as the Centre for Genetic Engineering and Biotechnology (CIGB) and BioCubaFarma enabled Cuba to achieve world-class biomedical innovation in vaccines, pharmaceuticals and preventive healthcare, despite severe economic constraints. Cuba demonstrates that high technological capability can exist without industrial capitalism at a mass scale.

However, Cuba’s model also faced limitations imposed by a US economic embargo, which for 66 years restricted access to global capital markets, limited industrial scale and the absence of a broad heavy industrial base. The US’ aim was regime change in Cuba, which has now intensified under the Trump administration, with threats of direct military intervention. The fact that Cuba has managed to survive until now is a testimony to the resilience of the Cuban revolution and its socialist system.

Together, Taiwan and Cuba represent two different responses to structural position within global capitalism. Taiwan pursued industrial upgrading through strategic integration into global production systems, while Cuba pursued scientific sovereignty and public-sector innovation under conditions of relative isolation.

The Philippines currently occupies neither position. It is neither industrially upgraded like Taiwan nor scientifically autonomous like Cuba. Instead, it remains integrated primarily into low-value service, assembly and extractive sectors within global supply chains.

Implications: Sovereignty eroded

The primary danger posed by Pax Silica is that infrastructure developed within this framework increasingly serves both economic and military purposes. Industrial policy becomes aligned with external security priorities, while technology transfer remains conditional and hierarchically controlled. Strategic sectors such as semiconductors, logistics, telecommunications and energy become integrated into systems whose highest levels of ownership, design capability and operational coordination remain external to the Philippines.

This integration also deepens long-term dependency. Once infrastructure, investment flows, export markets, military coordination and technological systems become integrated into US-aligned networks, disengagement becomes economically and politically costly. In effect, sovereignty is not formally abolished but progressively narrowed through layered economic, technological and military integration.

Pax Silica presents itself as modernisation and technological progress, but structurally it represents a securitised global production system organised around US strategic interests. For the Philippines, the decisive question is whether AI and high-level technologies will enable industrial sovereignty and social development, or whether they will reinforce the country’s role as a managed periphery within a global military-industrial order.

The experiences of Taiwan and Cuba demonstrate that technological development is possible under very different historical conditions. However, both cases show that technological advancement ultimately depends on state capacity, ownership of productive assets, control over technological systems, and strategic autonomy.

Without these foundations, AI and high-level technologies do not produce sovereignty. They produce incorporation and subordination.

AI’s role in this system

Within Pax Silica, AI functions primarily as a productivity and coordination layer within global supply chains. It supports industrial automation, logistics optimisation, surveillance systems, military planning and data management. AI is therefore inseparable from the broader geopolitical and military-industrial restructuring currently underway.

However, without domestic control over semiconductors, compute infrastructure, energy systems, and industrial design capacity, AI development remains dependent on foreign technological ecosystems. Under these conditions, AI becomes a mechanism of participation within externally controlled systems rather than a foundation for technological sovereignty. The Philippines may therefore contribute labour, infrastructure, minerals, logistics and low-to-mid-level technical functions, while remaining excluded from the highest-value and most strategic levels of technological production.

A socialist alternative

Beyond the immediate tasks of national-democratic development and industrialisation lies the broader socialist transition. Under capitalism, production is governed primarily by markets, profit maximisation and the treatment of labour power as a commodity. A socialist transition seeks to progressively subordinate these mechanisms to democratic social planning, collective ownership and production oriented toward human need rather than private accumulation — the conscious curtailment of the law of value as the central organising principle of economic and reproductive life.

Such a transition would require reversing privatisation in strategic sectors and reestablishing public ownership and democratic management over key areas of the economy, including energy, transport, telecommunications, finance, water and major infrastructure systems, as well as comprehensive agrarian reform. This would need to be combined with workers’ control and integrated national planning capable of coordinating industrial development, scientific advancement, ecological sustainability and social welfare. Rather than leaving investment decisions to private capital and global market pressures, economic priorities would increasingly be determined through democratic planning mechanisms rooted in socially necessary and useful priorities.

Within such a framework, HLTs and AI would no longer primarily serve corporate profitability and imperialist competition. Instead, they would be directed toward socially necessary production and long-term human development. Their primary functions would include strengthening public healthcare systems, improving disaster prediction and climate resilience, supporting universal education, renewable energy systems, modernising sustainable agriculture and food security systems, and coordinating infrastructure according to ecological and social priorities rather than private profit.

A socialist approach to AI and industrial development would therefore treat technology not as an autonomous force or commodity, but as part of a broader project of social transformation. Technological progress would be evaluated according to whether it expands democratic control over production, reduces social inequality, strengthens collective welfare, restores ecological balance, and deepens national and popular sovereignty.

The long-term objective is not merely industrial growth, but the transformation of the social relations that govern production and reproduction.

Reihana Mohideen is a National Council member of the Partido Lakas ng Masa (PLM, Party of the Laboring Masses) and the head of the party's international desk.

—--------

Reject Pax Silica and the Philippines’ transformation into a hub of imperialist war and militarised AI infrastructure

Partido Lakas ng Masa, May 20

The US-led Pax Silica initiative is not a project for genuine Philippine development. It was established by the US government in December 2025 as a “strategic initiative” — a geopolitical-industrial bloc designed to secure US and imperialist dominance over semiconductors, AI, critical minerals, energy systems and strategic infrastructure.

One of the main aims of Pax Silica is to contain China. Through the Enhanced Defense Cooperation Agreement (EDCA) and related military agreements, economic integration is being progressively embedded within US strategic and military architecture, tightening the Philippines’ entanglement in Washington’s regional confrontation with China.

Under Pax Silica, the Philippines is being integrated into a US-aligned production and military network, through industrial corridors, logistics systems, military agreements, telecommunications infrastructure, AI systems, and strategic energy programs based not on renewables but nuclear power development through a “civil-nuclear energy agenda.”

The country is being assigned the role of labour provider, mineral supplier, assembly platform, logistics corridor and strategic military outpost, while higher-value technological design, semiconductor control, compute infrastructure, and industrial command remain concentrated in the high-level technology-dominant capitalist states, such as the US, Japan, Australia, Britain, Israel and others.

This is not technological sovereignty. It is dependency. The Bongbong Marcos government presents Pax Silica as “modernisation” and “development,” but the reality is deeper imperialist penetration into the Philippine economy, infrastructure systems, energy networks and national development trajectory.

Ports, airports, logistics corridors, telecommunications systems and AI infrastructure are being developed as dual-use systems that integrate civilian and military functions. AI itself functions not merely as an economic tool, but as part of broader systems of logistics coordination, surveillance, predictive analytics, drone warfare and security management.

The integration of Israeli-linked firms and technologies into emerging AI and infrastructure systems is a particular concern given Israel’s role within the global military-technological complex and its deployment of advanced surveillance, targeting and warfare systems — all of which are being tested and deployed in the ongoing genocide against the Palestinian people.

Philippine infrastructure will thus become embedded within global systems of militarisation, surveillance capitalism and ongoing imperialist wars and occupations, including the US-Israel war against Iran and the genocide in Palestine.

The Philippines will remain structurally dependent on foreign capital and imported technology, with weak industrial foundations and oligarchic control over strategic sectors. Its lack of energy security and high vulnerability to climate shocks further compound these conditions, reinforcing patterns of uneven and externally dependent industrialisation.

Pax Silica does not resolve these contradictions. It will intensify them by transforming Philippine territory into a strategic military-industrial node within a US-led imperialist order.

The country is being increasingly positioned as a logistics and infrastructure hub for US military-strategic interests, a supplier of critical minerals for external industrial systems, a low-wage platform for AI-related data processing, a site for surveillance and dual-use technological systems, and a frontline state in the intensifying US–China rivalry.

A sovereign development path requires a fundamentally different orientation. A socialist and national industrial strategy would subordinate markets and private accumulation to democratic planning and social need through:

• public ownership of strategic sectors;
• comprehensive agrarian reform and food sovereignty;
• coordinated national state plans;
• democratic economic planning;
• sovereign and renewable energy systems;
• expansion of domestic research and development for science and high-level technological capacity; and
• workers’ participation in economic decision-making.

Within such a framework, AI and high-level technologies would be oriented toward healthcare, education, disaster resilience, climate adaptation, food security, sustainable agriculture, renewable energy transition and infrastructure planning — all guided by social need rather than militarisation or US geopolitical interests.

Technology must serve the people — not imperialist domination, oligarchic accumulation, surveillance, and war.

The struggle for technological self-determination is part of the struggle for national sovereignty, social justice, democratic control of the economy and socialism.

Reject Pax Silica!

Reject the Philippines’ transformation into a hub of imperialist war and militarised AI infrastructure!

For genuine national and socialist industrial development!

• 1

  Artificial Intelligence refers to the simulation of human intelligence by digital devices, performing tasks such as learning, reasoning, problem-solving and decision-making, through techniques such as machine learning, neural networks and natural language processing.