In July 2024 a routine cybersecurity update turned computers around the world into expensive paperweights. A botched software patch from CrowdStrike, a cybersecurity firm, crashed networks, impaired emergency responses, and grounded flights for days. The incident underscored how reliant modern societies have become on interconnected infrastructure systems.

Infrastructure shapes the basic operation of daily life, from sending emails to getting electricity. It represents a hidden foundation that connects supply chains, technology networks, and energy grids to make modern existence possible. These systems rely on subterranean hardwiring and ductwork, which scarcely warrant attention in normal times. But when networks fail – whether from a cyberattack, natural disaster, or damaged undersea cable, for example – societies suddenly notice their dependence on infrastructure.

Just ask Elon Musk, an overlord of today’s infrastructure politics. Musk’s control over critical networks gives him the power to affect how, and sometimes whether, Ukrainian forces counter-strike Russian troops and how US politicians communicate with social media followers. Musk’s influence has expanded from developing financial infrastructure – launching the online payment platform PayPal in his early career – to include satellite connectivity, space exploration, social media, electronic vehicles, automated driving technology, and US federal governance. A German regulator recently warned that Musk’s power had become so great as to be ‘state-like’.

But infrastructure power operates by different rules than traditional forms of state control. While military might or financial strength often operate according to linear calculations – more resources equal more influence – infrastructure power is inherently paradoxical. Greater control over networks does not necessarily translate into greater freedom to wield that control.

The key distinction lies in how infrastructure power binds those who wield it. Exercising influence through networks requires being deeply embedded within them. This creates a form of mutual dependency: the very connections that enable power also create vulnerabilities. The CrowdStrike incident of July 2024 demonstrates this vividly. A single software update from one company caused over $5 billion in losses and paralysed networks worldwide, not because of a hostile attack, but through the ordinary operation of interconnected systems. This new reality demands a rethinking of how power operates.

Empires have long used infrastructure to project influence and dominate far-flung populations and territories. The Roman Empire built roads and aqueduct networks to increase control over territories. The British Empire constructed railways to manage colonies, boost trade, and move troops. The United States’ ascent in the early 20th century also leveraged infrastructure building, through transcontinental railroads and the Panama Canal.

Today’s infrastructure politics extends these historical trends and contains several novel features, due to the way modern societies are hardwired. Modern networks rely on decentralised, algorithmic systems increasingly managed by private entities. These attributes create linkages and chains of dependency that transcend national borders and defy traditional governance mechanisms.

Before analysing these challenges, it is worth pausing on the ‘what’: what is ‘infrastructure’, exactly? The term connotes the stuff – the pipelines, cables, and transportation networks that provide access to clean water, energy, and other public goods. The word conjures up highways, ports, and railways, in part because of its etymology. French railroad engineers of the late 19th century used ‘infrastructure’ to describe the material and foundations that underlay railroad tracks, such as trestles, tunnels, and culverts. To engineers, ‘infrastructure’ described the material below that enabled the functionality above.

If that sounds vague, good, because its vagueness is part of what made the term popular in English. After the Second World War, NATO negotiators debated how to pay for the massive construction needed to rebuild western Europe. French representatives used ‘infrastructure’ to describe not just military bases, but radio towers, airports, and other construction projects required to prepare European allies to withstand communism.

Winston Churchill was among those who scoffed at ‘infrastructure’ as pretentious jargon used by ‘highbrows who are naturally anxious to impress British labour with the fact that they learned Latin at Winchester’. Some US politicians and journalists also mocked ‘infrastructure’ as abstruse ‘gobbledygook’. Despite the objections, ‘infrastructure’ caught on and has, in recent years, become fodder for white papers, stump speeches, and state-spending debates worldwide.

What differentiates today’s infrastructure from that of previous eras is its speed, ubiquity, and interdependence. Consider how communications networks have evolved. In the 19th century, telegraph networks relayed information across continents in ways that awed contemporaries. Today’s internet moves staggering volumes of data – hundreds of millions of terabytes every day. The difference between telegram networks and the modern internet isn’t just speed or volume. After all, streaming a 10-minute web video requires moving more data than a 19th-century telegraph office might have processed in a year, but comparing data traffic alone misses the bigger picture. Modern networks are fast, dense, and ubiquitous enough to enable new kinds of social relations and economic exchanges, such that today’s infrastructure has become an always-on, real-time foundation supporting nearly every aspect of modern life.

Infrastructure has become both the foundation of, and a frontier for, global power. What does this mean for today’s world order? Each era’s infrastructure creates new gatekeepers, power dynamics, and types of vulnerability. Understanding this evolution is essential, as the world is experiencing a transformation in how power operates through these systems.

The new geometry of global infrastructure power is perhaps clearest in the evolution of financial networks. As I argue in my book, Dollars and Dominion: US Bankers and the Making of a Superpower, the expansion of US banking networks overseas in the early 20th century provided an essential financial infrastructure for supporting the US’s rise as an international power. In the late 19th century, there was virtually no overseas network of US bank branches. In the early 1900s, several banks attempted to open offices in Asia, Latin America, and Europe, but their efforts bore little resemblance to a centralised, strategic offensive hatched by omnipotent Wall Street executives or DC imperialists. Instead, early US branching looked more like a set of bumbling false starts by New York elites seeking to capitalise on government connections.

The passage of the Federal Reserve Act in 1913 changed the nation’s financial system, including allowing new forms of overseas branching. The creation of a central bank network, combined with the global upheaval of the First World War, reshaped where and how US bankers could operate overseas. While many US branches overseas faltered in the early years, the infrastructure of bank branches proved resilient long-term, thanks in part to the structural support of the Federal Reserve System and favourable macroeconomic trends for US businesses following the war. When traditional financial centres like London retrenched amid fighting in Europe, US branches overseas could provide a foundation to expand the nation’s influence and support surging global demand for the dollar.

Modern infrastructure – and societies’ reliance on it to function – presents new challenges that reverberate in puzzling and important ways. In particular, the shadowy and distributed nature of today’s networks raises fundamental challenges for democratic governance and the stability of critical systems.

Consider the financial firm BlackRock, a juggernaut of modern infrastructure power. Amid the 2008 financial crisis, the Federal Reserve Board deputised BlackRock, the world’s largest asset manager, to oversee several relief programmes. This involvement was not a one-time crisis response; it cemented closer ties among central banks, investment firms, and private banks. The Fed’s reliance on BlackRock reinforced the firm’s centrality to the smooth operation of the global financial system. Since its crisis-related work, BlackRock has continued to expand, including the recent acquisition of the world’s third-largest infrastructure-finance firm, Global Infrastructure Partners.

BlackRock’s rise contains echoes of early 20th-century financial relations, such as the outsized influence of banking titan JP Morgan during the Panic of 1907. Amid the 1907 collapse, Morgan summoned key players to his Manhattan brownstone to avert meltdown. The gambit worked, but Morgan’s manoeuvring highlighted the danger of a system that concentrated so much power in a single individual.

BlackRock’s influence, by contrast, is diffuse, shadowy, and mediated through algorithms and partners. The firm manages more than $9 trillion in assets – an amount that exceeds the GDPs of Japan and Germany combined. More than size alone, it is BlackRock’s integration into the very fabric of global finance that represents its true power. The firm affects fields as diverse as retirement savings, social media, and climate politics. In 2020, Bloomberg declared it a ‘fourth branch of government’, but even that characterisation might understate its influence. After all, BlackRock also operates a leading software platform used by many of the world’s largest financial firms to evaluate and manage risk. Its impacts are so sprawling that asset management firms like BlackRock now affect, as scholars Benjamin Braun and Brett Christophers note, ‘how whole industries are organised and how countries insert themselves into the global monetary and financial system’.

Most critically, BlackRock’s power operates through systems that are opaque to public scrutiny and resistant to traditional forms of democratic oversight. The firm relies on complex algorithms, affiliates, and global digital networks. As financial decisions increasingly use AI and real-time data flows, the locus of power shifts away from corporate boards and individual decision-makers to lines of code, diffuse networks, and layers of intermediaries and subsidiaries. These new power-dynamics make it harder for the public and policymakers to understand and address potential abuses and systemic risks.

From the citizen-level perspective, the infrastructure machinations of asset management firms might sound remote and irrelevant to daily life, but the prominence of multinational financiers in planning, managing, and overseeing infrastructure matters. In today’s markets, investors increasingly turn infrastructure projects into financial commodities that can be bought and sold. Financial firms such as BlackRock depict infrastructure funds as offering resilience in a ‘high inflationary macroeconomic environment’. The funds might appeal to investors, but investors’ models provide little assurance that projects develop and operate according to communities’ needs on the ground. When financing moves up, decision-making becomes less connected to local communities, as legal scholar Nahuel Maisley has argued. The shift can subvert communities’ control over what gets built and how it operates.

As infrastructure becomes increasingly complex and entangled, unintended consequences and cascading risks become even harder to anticipate and manage. The recent example of US and EU-driven sanctions on Russia presents a stark example of the circuitous and unintended effects of infrastructure power politics. Following Russia’s invasion of Ukraine, US and allied policymakers sanctioned Russia’s central bank and other large financial institutions. One measure prohibited several large Russian banks from accessing SWIFT, the Brussels-based financial messaging platform, on which the world’s largest banks coordinate cross-border movements of money.

The sanctions laid bare the power of the United States and its allies to make the international movement of money a foreign-policy battleground. Previously, most countries trusted the efficiency and centralisation of a single global platform for settling financial transactions, but the recent sanctions exposed how vulnerable unaligned nations could become if similarly targeted for SWIFT exclusion. Russia and China had already developed their own cross-border payment platforms in 2014 and 2015, respectively, and those platforms saw sharp upticks in traffic after the sanctions. Recently, other nations have also explored and begun developing nationally controlled platforms.

The backlash highlights the nonlinear nature of infrastructure power. Yes, global networks like payment systems can be weaponised; doing so, however, can have unintended, counterproductive, effects that expose vulnerabiity. Wielding infrastructure power highlights operators’ dependence on the very infrastructure networks that provide that power in the first place, thereby increasing their investment in the status quo. This fact helps explain why US interests would suffer if others divested from SWIFT and why asset managers like BlackRock would have much to lose in dismantling the current geopolitical order.

Today’s interconnected networks are also more difficult to safeguard. Infrastructure systems, from urban sanitation to supply chains, have become increasingly entwined; their interdependence heightens the risk of cascading failures. ‘Single points of failure’ riddle the modern internet, such that its overall architecture is vulnerable to cascading outages, according to computing experts Barath Raghavan and Bruce Schneier. After all, the economic conditions that produced today’s internet rewarded lean budgets, outsourcing, and the maximisation of short-term profits. Companies trimmed overhead and embraced just-in-time delivery models to squeeze profits rather than investing in redundant back-up systems. These pressures created a brittle network dependent on hundreds of obscure companies performing a ‘small but essential role in keeping the internet running’, Raghavan and Schneier argue. Case in point: CrowdStrike, the failure of which triggered confounding and cascading problems worldwide.

Modern societies’ dependence on these networks raises questions about national security, global competitiveness, and social justice. Infrastructure is not simply a different theatre for waging traditional political competition; instead, it creates a new geometry wherein the interests of governments, corporations, and advocacy groups are channelled by, and filtered through, technological networks, far-flung legal jurisdictions, environmental constraints – and pure chance.

What does savvy, thoughtful leadership look like in this age of infrastructure? Tidy solutions are hard to come by, but one thing is clear. Relying on old models of diplomacy and competition is about as useful as trying to build nuclear weapons with an abacus and a hacksaw. What’s needed are new frameworks and governance models, which could take years to refine.

Modern infrastructure networks are fracturing along new faultlines, and leaders must prepare for a world of bifurcating infrastructure regimes, rather than universal standards. Already, the world has seen decades of integration begin to unravel, as nations increasingly assert control over critical networks. The shift is not unique to financial systems. Similar patterns have emerged in electric vehicle standards, telecommunications, and global shipping, driven largely by escalating tensions between the United States and China.

This fragmentation poses fundamental challenges for both governance and business operations. The effectiveness of international standards has always depended on members’ participation: the rules only work when key players agree to follow them. If major powers opt to build parallel systems rather than work within existing frameworks, the very idea of oversight becomes increasingly difficult to maintain. The result is not just competing systems, but competing visions of how infrastructure should be governed and managed.

A core assumption of modern business has been challenged: namely, that profitability comes from continuously expanding access to customers and their data. Companies built to access an ever-expanding pool of global customers now face a world of shrinking markets and digital barriers. This new reality demands more than minor adjustments. It requires fundamentally rethinking business models, technological architectures, and trading strategies. While adapting to fragmented networks may be painful, it will be essential for surviving a world where the push for universal connectivity can no longer be taken for granted.

A second challenge of today’s infrastructure politics involves attuning to the knock-on effects of sustaining complex networks, rather than fixating on upfront design and construction. Media coverage and the fanfare of ribbon-cutting ceremonies draw attention to the launch of new projects. While these early moments matter, they often pale in comparison to the slow accretion of influence associated with operating and maintaining infrastructure over the long term. That was certainly the lesson for inept US bankers fumbling overseas in the early 1900s. They didn’t have to be financial wizards; they just needed to be competent enough to both stave off bankruptcy and keep US government contracts, until the world changed with the First World War. Being in the right place at the right time was enough for their infrastructure power to pay long-term dividends.

China’s Belt and Road Initiative could provide a similar foundation for reaping long-term advantages. While often framed as a plan to develop the physical infrastructure for trade routes, the BRI’s lasting significance may involve second-order effects. As Chinese construction companies install broadband networks and develop ports worldwide, their affiliates and partners are poised to win maintenance deals and future construction contracts. Families of workers sometimes relocate and choose to stay long-term. Lending relationships can also create enduring ties, especially as the debt of some projects becomes increasingly denominated in Chinese yuan. These connections persist beyond initial construction and may reshape global power dynamics in ways not immediately apparent to project designers. Whether the networks are financial, digital, or mechanical, the most profound consequences often emerge slowly, reshaping geopolitical realities in subtle but far-reaching ways.

The key to surviving in an interconnected world is not preventing failure: it’s mastering recovery. In an age where institutions depend on sprawling networks of suppliers, automated systems and digital infrastructure, disruptions are inevitable. The question is how quickly organisations can bounce back. From solar storms to cyberattacks, the range of potential shocks and their cascading tendency defy traditional risk management. The true institutional superpower of our era may be the capacity to rebuild amid chaos.

Raghavan and Schneier point to a counterintuitive solution: institutions can practice failing. ‘Chaos engineering’ teaches organisations to simulate disruptions and rehearse the way systems and people rebuild. These controlled failures force both technology and teams to develop resilience, not by insulating an institution against shocks, but by learning to absorb and adapt to them.

This approach requires different implementations across institutions. Google’s chaos engineering will look nothing like the State Department’s or BlackRock’s scenario planning. But the core principle remains: organisations must build muscle memory for crisis response. This means breaking down internal silos, creating flexible response teams, and treating recovery as a core competency rather than protocols reserved for an emergency.

The world desperately needs new best practices around infrastructure development. Space exploration, arctic navigation, and artificial intelligence governance are among the recent challenges that exceed the capacity of existing regulatory models. What does good cooperation look like? What are the necessary and sufficient conditions for ensuring infrastructure networks advance social justice?

After all, interconnected networks can be a source of good: they can catalyse power-sharing and encourage diverse actors to collaborate around shared problems. The Covid-19 pandemic demonstrated how global health infrastructure could catalyse co-operation in developing and distributing vaccines. Infrastructure can provide the connective tissue that links diverse entities and pushes them to tackle shared threats, from climate change to future pandemics. Identifying what conditions enable and accelerate collaboration can improve the world’s preparedness for an infrastructure-dominated world.

The challenges are formidable, but so are the opportunities. By recognising infrastructure as the hidden foundation of world order, we can begin to develop the tools and strategies needed to navigate the complexities of our interconnected future. The major forces driving 21st-century change will come not just from the actions of states and corporations but from an intricate lattice of infrastructure networks that connects us all.