Ports, power grids, rail corridors, data centers and critical mineral supply chains are no longer just “projects.” They are the operating system of sovereignty. Infrastructure—networks that move energy, goods, and data—is the industry of industries. Whoever shapes it through contracts, standards, currency denomination and long-term maintenance, much of which is increasingly guided by data and systems powered by artificial intelligence (AI), will achieve lasting global influence.
Debates about “de-dollarization” often focus on reserve currencies. In the International Monetary Fund’s Monetary Composition of Official Foreign Exchange Reserves data, the US dollar accounted for approximately 57% of global reserves in 2025, with the euro a distant second. But official reserves are a lagging indicator. The most relevant change refers to infrastructure.
China soon recognized this. Between 2000 and 2023, it provided approximately $2.2 trillion in official loans and grants as part of its Belt and Road Initiative, much of which was invested in transportation and energy infrastructure. This model was never just focused on capital. By pooling financing, contractors, equipment and digital systems, China was exporting state capacity and integrating long-term dependencies. Projects like the Chancay megaport in Peru — which is majority owned by a Chinese operator and backed by billions in investments — illustrate how infrastructure can reconfigure trade routes and other dependencies. Similarly, the Addis Ababa–Djibouti Railway, financed largely by Chinese loans, dramatically reduced loading times between Ethiopia and the Red Sea.
The geopolitical implications of infrastructure investment are increasingly a priority for policymakers. The possibility of Chinese involvement in airport construction in Greenland raised security concerns in both Denmark and the United States. The new contest is not only between currencies, but also between competing infrastructure blocks.
For decades, American influence was based on military power, the dollar, and multilateral institutions. But while this architecture remains relevant, it is rapidly being complemented—and in some cases challenged—by infrastructure strategies.
Political tensions reflect this change. In 2024, US President-elect Donald Trump threatened to impose harsh tariffs on countries seeking alternatives to dollar-based billing and payments. At the same time, Western economies have expanded their own infrastructure initiatives. The G7 Global Infrastructure and Investment Partnership, for example, aims to mobilize $600 billion by 2027; the European Union’s Global Gateway commits up to €300 billion ($353 billion); and the Blue Dot Network (launched by Australia, Japan and the United States) seeks to certify high-quality infrastructure standards.
However, many countries perceive these efforts as slow and overly conditional. In a world facing climate shocks, demographic pressures and urgent development needs, the ability to deliver infrastructure quickly often outweighs governance concerns.
Various middle powers are redefining their own strategies accordingly. India, for example, is pursuing “corridor diplomacy” by supporting projects such as the Chabahar Port and the India–Middle East–Europe Economic Corridor. Instead of aligning exclusively with one bloc, it is taking advantage of infrastructures to cover itself, diversify and expand its own strategic autonomy.
There is also another critical change underway. Far from being limited to steel and concrete, infrastructure geopolitics increasingly extends to computing, data and AI. Corporate presentations reveal the magnitude of this transition. Tech companies such as Microsoft, Alphabet, Meta and Amazon are investing tens of billions of dollars annually in AI infrastructure, including data centers and specialized hardware. Its capital expenditures and associated depreciation now resemble those of traditional infrastructure sectors.
Semiconductor manufacturing has become a strategic choke point in this system. Facilities costing tens of billions of dollars anchor global supply chains and define access to advanced computing capabilities. But AI is not just another layer of infrastructure. It is the metainfrastructure that will shape how all other systems are planned, operated and optimized. If infrastructure defines geopolitical power, AI is increasingly defining infrastructure. It can improve network efficiency, extend the life of transportation networks, and enable more precise climate adaptation strategies.
But AI also introduces new forms of vulnerability. Remote control over optimization systems can function as a “kill switch” for critical infrastructure, and opaque or biased algorithms can systematically determine which regions or communities receive investment. In this context, infrastructure is no longer just about physical assets, but also about who controls the intelligence layer that governs them.
Nowhere are the stakes more visible than in Gaza. According to UN and World Bank assessments, by the end of 2025, approximately 90% of housing and infrastructure had been damaged or destroyed, and almost the entire population of 2.1 million people had been displaced. Reconstruction will require tens of billions of dollars, but without a lasting political settlement, the enclave’s new infrastructure could become an instrument of control rather than recovery. Infrastructure geopolitics is not inherently migratory. Transit corridors, energy systems, and housing can be designed to enable mobility and growth; but they can also be used to limit people.
The world is heading toward overlapping infrastructure ecosystems: a US-centric system based on open capital markets and legal enforcement; a China-centric system that combines state finance, contractors and integrated standards; and a diverse set of regional and middle power strategies.
The decisive question is not which system is the greatest, but which will ultimately be accepted as the default. Even when decisions about ports, energy systems and data networks are framed in purely technical or financial terms, there is an underlying political logic at play. These decisions are increasingly mediated by AI systems trained on historical data and designed to optimize efficiency, thereby narrowing the range of perceived alternatives.
George Orwell famously warned about controlling language. Today, we are witnessing an even more subtle form of power. Infrastructure, increasingly guided by algorithmic systems, risks making certain potential futures seem inevitable and others unthinkable.
The greatest danger is not that one power dominates, but that societies gradually lose the ability to choose between competing paths. The new world order is being built in concrete and coded in silicon. The ultimate sovereign act may not be to build or resist, but to recognize that these options remain open, before the systems we build begin to optimize us in return.
Bertrand Badré, former Director General of the World Bank, is Chairman of the Advisory Board of Project Syndicate, CEO and founder of Blue like an Orange Sustainable Capital. Saurabh Mishra, former director of the Human-Centered Artificial Intelligence Institute at Stanford University and economist at the World Bank and International Monetary Fund.