DOSSIER

DIGITAL LIFE

Big tech, military power, and the new digital-industrial complex

This is not just a curious piece of news. It is a sign of the times. Executives at the core of digital capitalism – leaders of companies that control platforms, data, and algorithms – are beginning to receive military patents and formally integrate into the United States Army. The announcement of Detachment 201 – Executive Innovation Corps, in 2025, makes this movement explicit: to bring together, within the military institution, the corporate culture of Big Tech and the routines of armed power1.

At first glance, the episode can be read as an administrative update. But the central point is another. When platform executives become part of the military structure – with patents and all – the relationship ceases to be merely contractual. The boundary between the civilian and military spheres becomes blurred. Innovation is not only applied to war: it becomes part of its very architecture.

In 1961, Eisenhower warned of the risk of the “acquisition of unjustified influence” by the military-industrial complex2. His target was the alliance between the arms industry, public contracts, and armed power. Today, this warning requires updating. What was once the arms industry and contracts has come to include cloud computing, artificial intelligence, digital surveillance, and global platforms. A military-industrial-digital complex is emerging, in which data and informational infrastructure become central components of strategic power.

This reconfiguration is based on a broader transformation: the conversion of social life into data. Zuboff described this process as surveillance capitalism, based on the extraction of behavioral signals and their transformation into predictive models capable of guiding conduct. Politically, the effect is decisive: platforms begin to operate as the cognitive infrastructure of society and, by extension, as an asset for intelligence, control, and coercion systems.

Snowden's revelations showed that the global digital communication infrastructure was already deeply penetrated by large-scale state surveillance programs, such as PRISM. The integration between civilian platforms and security apparatus does not depend on military symbols. It occurs at the structural level, in chains of technical cooperation and in zones of secrecy legitimized by “national security.”

In this context, contemporary warfare becomes less dependent on firepower and more on informational power. The battlefield is organized as an operational chain: detect, identify, track, prioritize, engage, evaluate – the so-called kill chain. When algorithms accelerate this chain, they shorten the time for human deliberation and shift critical decisions to opaque, often proprietary, systems. Violence becomes faster, more automatable, and less traceable in terms of accountability.

In 2024, reports revealed the use of the Lavender system, employed by Israel to select human targets in Gaza, through the automated cross-referencing of data and probabilistic classifications [5]. The central issue is not only the use of AI, but its logic: people converted into statistical categories; suspicion transformed into scores; prioritization done by models that are difficult to audit. In this arrangement, error ceases to be marginal and becomes structural.

If Lavender represents the top of the chain, Project Nimbus represents the base: the computing infrastructure. In 2021, Israel signed a billion-dollar contract with Google and Amazon for cloud services intended for the government and security forces.5 Without massive storage, database integration, and large-scale processing, algorithmic warfare cannot be sustained. When this infrastructure belongs to transnational corporations, state secrets and industrial secrets combine, making auditing and public control difficult.

In the United States, Project Maven provided a precedent. Created to apply AI to image analysis and drone surveillance, it involved Google and generated internal protests and resignations in 2018.6 Even when there is ethical resistance, the incentive structure remains. Furthermore, Maven highlights the dual-use nature of the technology: civilian systems easily converted into military applications.

Integration has been institutionalized since the last decade. In 2016, the Department of Defense introduced DIUx as a bridge between the military apparatus and the innovation ecosystem.7 Detachment 201 signals a step further: less bridging, more incorporation. The capabilities of digital capitalism cease to be peripheral and become part of the institutional engineering of the armed state.

After the owners of big tech companies were given prominent positions at Donald Trump's inauguration, the heads of Meta (owner of Facebook, Instagram, and WhatsApp), Apple, Microsoft, and OpenAI had dinner with the US president at the White House in early September. "Thank you for your incredible leadership, including bringing this group together," said Bill Gates, the billionaire owner of Microsoft, to President Donald Trump.

"Big tech companies are geopolitical machines. Let's not kid ourselves. Technology is not just a means to an end. Technology is one of the main instruments of global political, economic, and military power. Trump himself says that big tech companies are the front line of American power," explained Sérgio Amadeu da Silveira—sociologist and professor at UFABC.

The Military-Digital Complex and the New World (Dis)order...At the dawn of the internet, digitalisation was heralded as the key to unlocking the emancipatory virtues of the free market — spreading knowledge and economic opportunity, but above all ensuring peace and strengthening democracy. Today, it seems instead to be reviving old contradictions. Digitalisation has not only revolutionised how we communicate, produce, and consume; it has also fostered an unprecedented concentration of economic and technological power. Consider the market capitalisation, revenues, and profits of the US-based Big Tech firms in 2024 and 2025. In March 2025, their combined market capitalisation was three times the GDP of Germany and not far from that of the entire Euro Area ($16 trillion). In 2024, their share of profits over revenues was at 27 percent, a very high value for US companies. R&D expenditure was 13 percent of revenue.

This concentration of techno-economic power breathes new life into the theses of thinkers like Hobson and Lenin, who revealed the imperialist nature of capitalism by linking war to the expansionist strategies of the great industrial monopolies of the early twentieth century. Old contradictions — inequality, instability, and the fractures within political and institutional systems that find in war their ‘natural’ outlet — now wear a new technological mask. The clash is between two military-digital complexes, the United States and China, locked in an increasingly violent struggle for control of markets, technologies, and critical raw materials. The digital sphere has become their privileged battleground: a vast panopticon where the profit-maximising strategies of digital oligopolies (which depend on constant surveillance and the extraction of data from those — ourselves included — who rely on their services) converge with the security, geopolitical, and military objectives of their respective states.

It is a perverse alliance. Private capital monopolises infrastructures (data centres, undersea cables), technologies (cloud and AI), and knowledge — codified in the patents they accumulate or embodied tacitly within organisations, and thus inaccessible to outsiders — now indispensable for conducting virtually any social or economic activity.

The state facilitates this process and seldom resists it (though tensions and contradictions abound), caught as it is in a relationship of mutual dependency. It cannot do without the technological and infrastructural capacities of Big Tech; without them, many of its objectives — both civilian and military — would be unattainable. Nor is it eager to curb the economic power of those who control the (social) platforms where public opinion and political consensus are shaped.

Through their respective Big Tech firms, the US and Chinese governments can maintain other nations within their spheres of digital subordination — possessing ‘eyes and ears’ that deliver a constant and invaluable stream of information.

Yet, dependence runs in the opposite direction as well. For Big Tech, cultivating a stable alliance with the state is not optional — it is a matter of survival. Their profits depend on their ability to monopolise network infrastructures and the data flowing through them. Hostile regulation or moves to bring these infrastructures under state control could severely limit, or even destroy, their capacity for accumulation. The same would be true of any serious increase in taxation.

And if the global economy slows — crippled by commercial, technological, and military wars, and by pervasive uncertainty — then the state, and particularly military spending, becomes an essential lifeline for preserving profit margins.

War, moreover, offers technological opportunity. It channels massive funding into military research in fields where Big Tech already holds dominance — automated command and control systems, artificial intelligence, and autonomous weapons. Active participation in conflicts also provides an unparalleled testing ground, where new applications can be refined under extreme conditions, free from oversight or ethical constraint.

This transformation is not limited to major powers. It would be a mistake to treat this scenario as something distant. Brazil is already experiencing, on its own scale, the advancement of surveillance technologies associated with urban security. Totems and cameras are spreading throughout large cities. In São Paulo, Smart Sampa is presented as a modernization of monitoring with facial recognition and artificial intelligence.8 The problem is not the stated objective of combating crime, but the infrastructure itself: once installed, it tends to expand, accumulate data, and migrate to a different purpose, especially in contexts of low governance and weak independent auditing.

The risk increases when these technologies are politically captured. The case of the “parallel ABIN” serves as a warning: intelligence apparatuses can be instrumentalized to monitor opponents, civil servants, journalists, and authorities.9 With digitization and automation, reach and capacity grow rapidly.

There is also the dimension of connectivity. Public records indicate the use of the Starlink satellite network in a military context in Brazil.10 The operational gain is real in remote areas. But the strategic precedent is unequivocal: critical communication infrastructure can become dependent on a foreign private provider, subject to corporate decisions and external pressures.

According to a consolidation of public data11, the regional distribution of the total of approximately 313,700 Starlink accesses in Brazil, in November 2024, reveals a strong concentration in the North (~30.7%) and Southeast (~29.5%) regions, followed by the Central-West (~20.4%), South (~10.9%) and Northeast (~5.3%). A hybrid pattern: high penetration in remote areas and significant presence in economically central regions.

This contrast helps to understand why some central states have begun to explicitly politicize technological dependence. In 2026, France prohibited public servants from using American videoconferencing tools such as Zoom, Google Meet, and Microsoft Teams, imposing migration to a state platform developed by its interministerial digital authority and hosted on national infrastructure. The measure is not administrative: it is a gesture of sovereignty. This is part of a deliberate movement to reduce structural dependence on services based in the United States and to reclaim control of the basic layers of communication by the state.

The same horizon guides the social regulation of platforms. French parliamentarians approved a bill prohibiting the use of social networks by minors under 15 and expanding restrictions on the use of cell phones in schools. The political message is clear: platforms are not treated merely as companies, but as infrastructures capable of affecting subjective formation, the public sphere, and democracy.

Recent episodes make this dimension politically unavoidable. In Iran, during near-total internet blackouts, Starlink terminals were used clandestinely to circumvent the state blockade and maintain external communication. A private infrastructure, controlled by a corporation based outside the country, came to play a functionally decisive role in a context of internal instability. Systems of this type can acquire direct strategic weight.

For countries like Brazil, this international experience serves as a warning. By incorporating similar solutions into military, intelligence, or public security communications, the State assumes a structural dependence on infrastructure over which it does not have full sovereign control. In scenarios of institutional crisis, conflict, or international pressure, this dependence is not neutral: it can become a strategic vulnerability, not only due to a deliberate decision by the provider, but also due to the existence of an external point capable of conditioning, degrading, or interrupting critical communications. In this sense, connectivity ceases to be a technical issue and becomes part of the core of sovereignty and national security problems.

This set of issues reveals a frequently underestimated weakness: the Brazilian State's systemic dependence on proprietary software, closed platforms, and poorly auditable networks. When sensitive communications and government operations rely on ecosystems controlled by transnational corporations, sovereignty ceases to be merely a geopolitical concept and becomes an operational condition. Infrastructure dependence is power, as it defines the limits of what is possible even before the political decision.

This is where the idea of ​​a tutelary democracy gains traction. Elections and democratic rites are maintained, but the center of gravity of strategic decisions shifts to a hybrid, technocratic, and corporate sphere, protected by algorithmic secrecy and opacity. The language of efficiency tends to naturalize political choices as if they were technical. But neutrality is impossible when systems define suspects, prioritize threats, select targets, or expand surveillance.

Eisenhower did not foresee algorithms, but he anticipated the mechanism: a circuit of interests fueled by war, fear, and exceptionality.12 The new complex produces not only weapons but also operational categories: risk, suspicion, priority, target. When these categories become code, democratic control loses its capacity for intervention: it no longer debates means and ends transparently; it begins to debate outputs, effects, and consequences.

In the end, the decisive point is not technological, but political. The expansion of surveillance, automation, and data integration systems – increasingly dependent on private infrastructure and opaque platforms – shifts the center of gravity of strategic decisions to opaque circuits where corporate interests and military logics converge. For democratic societies, this creates a new challenge: it is not enough to control the use of force; it is also necessary to control the informational means that make this force possible and operational. Without clear rules of governance, auditing, transparency, and technological sovereignty, the risk is that democracy will remain formally intact, but materially limited by infrastructures that it does not fully understand, control, or oversee.

by: Celso P. de Melo is a retired Full Professor at UFPE, a CNPq 1A Researcher, and a member of the Brazilian Academy of Sciences.