Digital Infra: The Coming Reckoning
The internet was built for research papers and information sharing. Now we’re trying to run $10 trillion in global trade on it. It’s breaking.
Overview
Over the last three decades, digital economy has moved from the periphery to an increasingly center stage of the overall global trade. It started with the dot-com era, and has evolved to stablecoins and digital currencies emerging as the preferred settlement mechanism for global trade. Today, as per the World Bank, the digital economy is about 15% of the overall economy but it’s growing at almost 2X the overall economic growth rate.
The second-order impacts of this increasing share range from the individual to the country-level policy. It is changing the design, distribution, and consumption of products and services, the communication channels and trends between firms and consumers, the payment and financial services layers and rails, and even the policies each country is enacting to participate in this growing digital economy while protecting its constituents.
However, much of the underlying infrastructure that is supporting this growing share of the digital economy is a patchwork of many disparate and often ill-fitting systems designed as an afterthought to plug certain gaps on top of the foundation of this economy - the internet. What we fail to perhaps often acknowledge and appreciate is that the foundation of the internet was built more for information sharing rather than trade, and thus has its limitations in carrying the increasing weight of the digital economy.
The development of the digital infrastructure closely mirrors the expansion of the railroads and the telegraph system. Initially, railroad construction in the 1820s and 1830s resulted in a fractured web of independently owned, localized lines with incompatible physical attributes, such as varying track gauges. It was not until the period between 1860 and 1890 that administrative practices and technological aspects were unified, allowing for a genuinely integrated national and global transportation system. The digital economy is currently in a similar phase of growth and fragmentation leading to what could perhaps be national unified and interoperable digital infra.
Below are the signals across 5 dimensions where change seems to be emerging from:
The Technological Question
The Geopolitical Aspect
The Macro Economic Aspect
The Societal Perspective
The Energy & Security Perspective
The Technological Question: Control, Trust and Security For Trade
The internet was conceived as an open and information-sharing network. Therefore, much of it is centered around ‘information’. The proliferation of digital information gave rise to websites, websites gave rise to more types of information sharing such as music and video, this exchange led to social communities and commerce, and these gave rise to digital payments and advertising. The introduction of smartphones somewhere in the middle of this journey allowed for all of this to be accessible on the move.
The digital economy was thus centered around ‘information technology’ and built infrastructure vertically - new platforms on existing architectures and protocols/standards, new services leveraging old networks, innovation compounding on foundations laid years ago.
But the locus of the digital ecosystem has shifted from ‘information’ to ‘trade’ and that shift requires a shift in the tenets of underlying infra as well - a move towards ‘trust’, ‘safety’ and ‘resilience’ as its underlying foundations - akin to the sea shipping lanes for the physical economy.
Many issues we see with the digital economy today - fraud, cybersecurity, surveillance, monopolization, critical infra risks - all ladder up to the market players exploiting the architecture and market dynamics (such as monopoly networks, the anonymity offered by the internet, sharing of data to access websites, etc.) which were fair for information exchange but are detrimental to fair and secure trade.
Three themes capture the emerging infrastructure signals:
Concentration of Control & Toll-Booth Economy: The current server-client architecture of the internet stems from the information-sharing era. In the early days of the internet, the compute power of devices was limited and information was concentrated in research labs, academic institutions, corporates, etc. This gave rise to a client-server architecture where the consumer needed to send their request and info to the server to consume any product or service. As we evolved digital products & services from information to networks, entertainment, commerce, and critical services - this architecture remained, creating large pots of data & information all within the silos of the companies managing these platforms and services. Often, these companies themselves are concentrated within certain geographical regions - creating a critical dependency, surveillance, risks of being cut off as well as an imbalance of trade. Now, with blockchain and tokenized economies, new architectures are developing.
Lack of Identity, Trust and Legal frameworks: Fair and secure trade requires ‘trust’ in the transaction. However, this infrastructure layer lags behind the global accessibility that the digital economy is affording. Trust means being able to verify the identity of the parties involved, to be able to judge the authenticity of a transaction, to have secure and unrestricted means of payment, and legal protection in case of a transaction going south. There is no centralized mechanism for a buyer on a website to authenticate the legitimacy of the business. Sending payments globally is still cumbersome and expensive for many businesses, and increasingly complex due to fraud prevention and AML checks etc. The anonymous and digital nature of the internet from the information-sharing era has also evolved into the ‘trade’ era - with minimal support and a lack of legal recourse across global jurisdictions in case of a ‘bad trade’. As of recent data, 850 million people globally lack any form of official identification, and a staggering 5 billion people live in jurisdictions without a secure digital ID capable of facilitating secure online access to public and private sector services. Furthermore, only 96 economies have fully operationalized the legal frameworks and technological infrastructure necessary for e-signatures, a fundamental requirement for digital trade.
Security Vulnerabilities: Core protocols that direct traffic, such as DNS, were designed for an era of high-trust academic networks. Simultaneously, the progress in quantum computing is opening the risks of the current security and cryptographic protocols being obsolete. These encryption algorithms (such as RSA and ECC) currently secure the global financial system, digital trade, and private communications. Adversaries are already employing “harvest now, decrypt later” strategies, intercepting highly sensitive encrypted data today with the intent to decode it when quantum capacity matures. The structural gap here is quite risky: upgrading the entirety of the internet’s cryptographic foundation—including web browsers, VPNs, firmware signatures, and legacy rails—requires a level of coordinated execution, discovery, and automation that current fragmented governance models are ill-equipped to handle.
This is leading the countries to invest in digital public infrastructure, even though the pace of its development lags the growth of digital economy especially with the acceleration afforded by AI.
The Geopolitical Aspect: Digital Sovereignty In an Increasing Multipolar World
Countries are increasingly re-considering their control and design of digital infrastructure. What was once global, open, and commercial - building data centers, laying fiber optic cables, processing payments - is now a strategic national capability. No country wants to be dependent on systems and infrastructure offered by another - and run the risk of being held hostage.
The biggest challenge in the infrastructure for governments here is increasing concentration and dependency on foreign commerce, and payments platforms, and the AI and cloud infrastructure.
Canada has launched a sovereign cloud infrastructure and mandates that data generated within its borders must be copied to it. India requires payment transaction data to remain on domestic servers. The European Union is increasingly redefining digital rules, and some countries have banned digital products made elsewhere for govt work, due to safety concerns.
AI has increased the urgency of conversations around digital sovereignty in policy making. The US has introduced a regulation to vet AI models, and Anthropic restricted access of its newest model to hand-picked companies, most of which were concentrated geographically in North America.
Regulators and governments globally are watching these developments and increasingly looking to be self-reliant and de-risk their digital economies. Additionally, the digital sovereignty play is not just limited to software but also extending into hardware.
China processes 60% of the world’s rare earth elements and controls 80% of global refining capacity, giving it leverage over the physical components that underpin all digital infrastructure. The semiconductor supply chain, previously optimized across global networks with design in the US, fabrication in Taiwan, and assembly in Southeast Asia, is now being duplicated regionally at significant cost. The US CHIPS Act commits $52 billion to domestic semiconductor manufacturing explicitly to reduce dependence on Asian supply chains. More recently, modems, AI models, and auto parts have also come under regulation to prevent ceding sovereign control over critical infra.
The implications of this are both economic and technological.
On one hand, this would mean significant CAPEX investments into building out the digital infra which will push money out of financial markets, thus deleveraging asset valuations. On the other hand, the technical architecture of digital services is being reengineered. We are moving from an architecture for global distribution and optimization to regional restrictions and perhaps even on-premise and edge computing.
A fintech startup in 2015 could build once and deploy globally. A fintech startup in 2026 must now build for different data storage requirements, regulatory frameworks, payment rails, and infrastructure partners in each major market. Development costs rise when similar solutions must be rebuilt for each jurisdiction. Innovation slows when platforms cannot leverage unified datasets to train algorithms or detect fraud patterns globally.
While this preserves digital sovereignty, the efficiency gains and technological innovation that propelled digital economy growth are being partially reversed. It is the digital equivalent of Autarky - where closing the borders leads to self-reliance but less growth and increased inflation.
The Macro Economic Aspect: Tariff Era and End of Open Digital Trade
Structurally, the global digital economy has relied on a foundational agreement to keep electronic transmissions free from taxation. However, this architectural pillar collapsed at the World Trade Organization’s (WTO) 14th Ministerial Conference (MC14) in March 2026, where members failed to reach a consensus to extend the moratorium on customs duties on electronic transmissions. The expiration of this moratorium, which had been in place since 1998 and protected digital trade from punitive tariffs, fundamentally alters the economic infrastructure of the web.
With increased barriers, protectionism, and tariffs in global physical trade, it is only a matter of time before that perspective shifts into digital trade. The fundamental aspect here is that the global barrier-less architecture of Internet allows firms based in one geography to extract value from the consumers and data in another country. Many of the most dense clusters of data generation are in emerging economies, but they have no share in this economic value created and is leading to transfer of wealth to developed economies which tend to be, the current providers of digital platforms.
Countries such as India, Brazil, and South Africa actively opposed making the tariff ban permanent and pushed for its expiration. The argument being that developing countries could generate forty times more tariff revenue by imposing customs duties on electronic transmissions than developed nations can. On the other hand, 23 countries—including the United States, the United Kingdom, Japan, and Mexico—recently signed an agreement committing not to impose e-commerce customs duties among themselves. Thus, block trade policies are emerging in digital trade as well.
A number of European countries have implemented active Digital Service Taxes, including Austria, France, Hungary, Italy, Spain, Turkey, and the United Kingdom.
A software company that once distributed globally at near-zero marginal cost would face tariff calculations, customs procedures, and compliance costs at each border. Cloud providers would need to factor in duties on cross-border data processing services.
The loss of the moratorium and lack of progress on the OECD framework for digital taxations highlights a profound misalignment between the inherently borderless design of digital infrastructure and the increasingly protectionist impulses of sovereign governments, setting the stage for a fractured, regionalized internet where data flows are regulated identically to physical cargo.
The Societal Perspective: Consumer Fatigue and the Extraction Model
The business model of the digital economy has been digital consumption and extraction. Platforms extract attention, convert it to data, monetize that data through targeted advertising or algorithmic optimization, and reinvest proceeds into acquiring more users to extract more data. Much like lending and credit drives the physical economy. This worked when consumers perceived net benefits - convenience, connectivity, entertainment - that exceeded the costs of surveillance and data harvesting.
This is where increasing consumer fatigue, social unrest, and pressure on regulators is shifting the equation.
The EU Digital Services Act (DSA) requires platforms to mitigate “systemic risks” related to addictive design and algorithmic rabbit holes. Australia has banned social media for teenagers. The EU Digital Markets Act (DMA) prohibits platforms from combining personal data collected across their different services (e.g., mixing search data with social media data) without explicit user consent.
Consumers also increasingly understand that their data generates value they don’t capture. The asymmetry is becoming visible - creating pressure on the regulators. Countries such as Australia have legislated a consumer data right providing consumers the right to own, control, and share their data across key sectors such as financial services, energy, and health.
Recently, a book titled “ The Anxious Generation” chronicling the psychological impacts of high screen time on youth by Jonathan Haidt spent 100 weeks after its launch on the New York Times bestseller list.
This emerging shift raises two questions that are tied to the infrastructure of the digital economy:
Firstly, who owns the data and whether individuals should receive compensation for data that generates economic value. Proposals for “data dividends” or “national data funds” suggest mechanisms to distribute value back to data contributors, fundamentally altering the architecture and digital business models built on free data extraction.
Secondly, surveillance fatigue manifests in changing consumption patterns. Younger demographics increasingly avoid platforms perceived as extractive, seeking alternatives that promise privacy or user control. The rise of decentralized platforms, encrypted messaging, VPN usage, Ad blockers, etc. means consumers are actively resisting the extraction model.
This resistance creates infrastructure implications. If platforms cannot extract data as freely, they cannot train algorithms as effectively. Recommendation systems degrade. Targeting becomes less precise. The efficiency advantages that made digital platforms dominant - their ability to match supply and demand, personalize experiences, optimize operations - depend on data flows that consumers are increasingly willing to block.
This societal resistance to unrestricted data extraction directly intersects with the mounting pressures on the physical resources required to sustain these current models.
The Energy & Security Perspective: Critical Infrastructure Under Pressure
The physical infrastructure supporting digital services faces constraints that compound technology, economic, and geopolitical considerations.
Data centers require enormous amounts of power - a single hyperscale facility consumes enough electricity to power a small city. McKinsey projects 156 GW of AI-specific data center capacity demand by 2030, requiring 125 GW of incremental additions. Goldman Sachs forecasts data center power demand increasing 165% by 2030 relative to 2023 levels. The first constraint thus is power availability.
Water poses another constraint. A single hyperscale data center can consume up to 1.5 million liters per day for evaporative cooling. As high-density AI racks dissipate 40-60 kilowatts of power - eight times traditional enterprise levels - cooling requirements intensify.
While currently there is a massive influx of CAPEX to build these energy and water infra out, the sustainability of the energy required per unit of digital economic growth remains questionable.
The physical backbone of the digital economy - the submarine fiber-optic cables - have become highly politicized. Over 95 percent of global internet traffic, facilitating roughly $10 trillion in daily financial transactions, travels through these networks. In regions like the Pacific Islands, which rely on these cables for basic connectivity and economic participation, infrastructure investment has become a proxy for geopolitical dominance.
The United States, Australia, and Japan are actively coordinating to fund undersea cables in the Indo-Pacific. Their explicit strategic aim is to crowd out Chinese-owned infrastructure, thereby reducing Beijing’s perceived espionage capabilities and maintaining leverage over global data flows. China, conversely, leverages its vast capital and state-backed telecommunications firms to offer rapid, cost-effective digital connectivity to developing nations, accompanied by broader strategic alignments. China’s recent live-fire naval drills in the Tasman Sea and the unveiling of deep-sea cable cutters serve as unmistakable demonstrations of power, highlighting the vulnerability of these underwater choke points.
Lastly, cybersecurity threats compound infrastructure vulnerabilities. Critical infrastructure - power grids, water systems, financial networks - increasingly operates through digital controls that present attack surfaces. Nation-states develop cyber capabilities to disrupt adversary infrastructure. Criminal networks target financial systems, healthcare providers, and municipal governments with ransomware. The infrastructure that enables digital services is itself vulnerable to digital attack, creating cascading failure risks where compromising one system can disable others dependent on it.
Implications: The Infrastructure Realignment
All of these points lead back to one question: Do we need to rethink the infrastructure on which we are increasingly expanding the digital economy footprint?
While the fundamental transport protocols provide a baseline of connectivity, the higher-level infrastructure—identity frameworks, data sovereignty standards, and cloud computing ecosystems—remains deeply siloed and heavily monopolized.
Larger countries with economic and technical capacity are building comprehensive digital infrastructure stacks and supporting physical supply chains. Smaller countries face stark choices. They cannot afford to build comprehensive digital infrastructure stacks. They must choose whose infrastructure to depend on. This choice carries strategic implications.
This represents a fundamental shift in how digital infrastructure operates and who controls it. The open internet built on voluntary cooperation and shared protocols is giving way to managed digital ecosystems with defined borders, access controls, and sovereignty requirements. The efficiency gains that drove digital economy growth - instant global reach, near-zero marginal costs, network effects at planetary scale - are being traded for resilience, sovereignty, and control.
Whether existing digital infrastructure can withstand coming changes depends on how rapidly and completely this transition occurs. A gradual shift allows infrastructure to adapt - data centers get built in new locations, supply chains diversify, platforms adjust to regional requirements. A rapid shift triggered by geopolitical conflict, major cyberattacks, or infrastructure failures could overwhelm adaptation capacity, leaving critical gaps in digital infrastructure precisely when AI and other emerging technologies demand more.
The digital economy is no longer building vertically on stable foundations. It is simultaneously building new foundations while the old ones fracture beneath it, hoping the new infrastructure completes before the old infrastructure fails. To survive this realignment, organizations must transition their strategies from efficiency-optimized global models to resilience-optimized regional architectures.