By Lina Irawan 
It may sound improbable, a relic of a less interconnected age, yet recent events suggest otherwise, offering stark warnings about the vulnerabilities inherent in our global digital infrastructure. The year 2024 saw significant subsea cable damage that caused widespread internet outages, crippling banks, fintech platforms, and cloud-based businesses across large swathes of Africa. The economic ripple effects were immediate, disrupting critical services and highlighting the continent’s growing reliance on digital connectivity. Following swiftly in 2025, a series of severe Red Sea cable cuts further compounded these disruptions, profoundly impacting internet access across Asia and the Middle East. These incidents weren’t isolated anomalies; they forced internet traffic to be rerouted through significantly longer and less efficient paths, dramatically increasing latency for essential cloud services and imposing substantial additional operational costs on businesses reliant on real-time digital operations. Such disruptions can cost economies billions, not just in direct losses but in lost productivity, damaged trust, and deferred investment.
These incidents serve as a potent and urgent reminder that what we colloquially refer to as “the cloud” is neither purely virtual nor inherently invincible. We often conjure an image of data floating intangibly somewhere above us, a ubiquitous, resilient network impervious to physical harm. In reality, the vast majority of our digital universe, the very backbone of global communication, lies not in the sky but at the bottom of the ocean. More than 99% of all international data traffic is carried, not by satellites, which play a crucial but niche role, but by an intricate, physical network of thousands of subsea fiber optic cables. These silent arteries, spanning hundreds of thousands of kilometers across the world’s oceans, are the unsung heroes of the internet, enabling everything from global financial markets to instant messaging. Without them, the global digital economy would grind to a halt.
The uninterrupted functioning of the regional digital economy, particularly in dynamic growth markets like Asia, is predicated on continuous and heavy investment in critical transit hubs. Countries such as India and Singapore have emerged as indispensable nodes in this global network, not by accident, but through strategic foresight and sustained capital injection into their digital infrastructure. These hubs serve as vital intersections where numerous subsea cables make landfall, connecting continents and facilitating the immense flow of data that powers modern commerce and communication.
However, the advent of Artificial Intelligence (AI) is placing unprecedented and escalating demands on this foundational "cloud" infrastructure. AI, poised to be the primary engine of today’s digital economy, requires computational power and data transfer capabilities far beyond previous generations of technology. AI workloads, encompassing everything from training massive language models to real-time inference and complex data analytics, necessitate ever-greater bandwidth. What was once considered cutting-edge, 400 gigabits per second (Gbps), is now becoming the baseline requirement for hyperscale data centers, with 800 Gbps already emerging as the next standard. This exponential increase in demand is driven by the sheer volume of data processed, the complexity of AI algorithms, and the need for instantaneous communication between distributed computing resources.
As countries and businesses globally race to scale AI capabilities, much of the strategic focus and investment has understandably been directed towards compute power—the specialized processors (GPUs, TPUs) and the massive, energy-intensive data centers that house them. Yet, a critical question arises for regions like Asia, which are at the forefront of AI adoption: Is the region inadvertently overlooking a key AI infrastructure bottleneck? The powerful networks connecting these distributed compute systems, particularly the subsea cables, must keep pace with this explosive growth, or the entire AI ecosystem risks being severely hampered.
Without sufficient, proactive investment in robust subsea cables and integrated end-to-end networks, businesses could find themselves in a precarious position. They might spend heavily on cutting-edge AI infrastructure—the latest chips, advanced cooling systems, and state-of-the-art data centers—only to see their potential returns eroded by persistent disruptions, higher operating costs due to inefficient data routing, and unreliable access to the massive datasets that feed AI models. This creates a critical paradox: substantial investment in the visible components of AI without commensurate investment in the invisible arteries that connect them.
For the C-suite, this scenario presents a looming crisis, a strategic disconnect highlighted by recent industry research. A comprehensive study by Tata Communications revealed a telling contradiction: while AI investment is recognized as a board-level strategic priority for a significant 77% of enterprise leaders, a staggering 65% of these enterprises continue to operate on legacy or still-developing infrastructure. This disparity suggests a dangerous gap between aspiration and reality, where the ambition for AI-driven transformation is not matched by the foundational investments required to support it. The short-term focus on compute capacity often overshadows the long-term strategic necessity of robust connectivity, which, while less glamorous, is equally vital.
Compounding this challenge is the nature of AI workloads themselves, which are making global traffic patterns increasingly unpredictable and dynamic. Unlike traditional enterprise applications with relatively stable data flows, AI workloads can generate sudden, massive spikes in demand as models are trained, updated, or deployed across distributed environments. To effectively manage this volatility and protect both the bottom line and operational stability, enterprises require dynamic, flexible data-center-to-data-center connectivity that can scale rapidly and intelligently with demand. Meeting this evolving need is not a static goal but an ongoing commitment that will require continuous and substantial investment from network providers in both the physical infrastructure and the advanced capabilities (like Software-Defined Networking and network automation) needed to deliver scalable, resilient, and agile connectivity.
Singapore’s ambitious pursuit of AI leadership is often framed around critical pillars such as compute capacity, cultivating a skilled talent pool, and establishing clear regulatory frameworks. While these elements are undeniably important for fostering an AI-ready ecosystem, discussions frequently overlook a fundamental vulnerability: what happens when the vast quantities of data that AI systems depend on cannot be transmitted quickly, reliably, or securely across borders? An AI model, no matter how sophisticated, is only as effective as its access to data, and data access is intrinsically linked to network performance.
In this context, Singapore and India are solidifying their positions as two of Asia’s most pivotal data transit hubs. Their emergence is underpinned by a combination of strategic geographical locations and rapidly expanding digital infrastructure. Singapore, with its unparalleled political stability and robust regulatory environment, serves as a crucial gateway to Southeast Asia, boasting one of the world’s densest concentrations of hyperscale data centers. This concentration makes it an ideal point of interconnection for regional and international data traffic. Meanwhile, India, with its immense and growing digital economy, acts as a critical link between Asia, the Middle East, and Europe. Supported by multiple major cable landing points along its vast coastline and rapidly expanding data center capacity in burgeoning tech cities like Mumbai and Chennai, India provides essential redundancy and connectivity options. Together, these two nations represent indispensable hubs for international data traffic, underscoring the urgent need for resilient, diverse, and high-capacity connectivity routes to ensure data moves efficiently and securely between major financial centers, commercial workloads, and vast user bases across continents.
To truly address the challenges posed by AI, the industry must fundamentally rethink its approach to network architecture. For decades, infrastructure was designed and built to support predictable, relatively linear traffic patterns characteristic of traditional internet usage. Today, however, infrastructure must evolve to accommodate the increasingly volatile, bursty, and massive data demands driven by AI. This paradigm shift necessitates a more holistic and integrated approach, fostering seamless connectivity between subsea cables, terrestrial fiber optic networks, and individual data centers. A bottleneck at a coastal landing station, for instance, can immediately negate the immense speed and capacity of a state-of-the-art transoceanic cable, creating a single point of failure that ripples through the entire network.
However, ensuring these critical cables remain secure and operational requires far more than merely laying fiber in the ocean. True network resilience demands a coordinated and multi-faceted focus across four crucial dimensions: meticulous capacity planning to anticipate future demand, robust network diversity through multiple routes and providers, proactive protection and rapid repair readiness, and disciplined lifecycle operations that ensure long-term stability.
This behind-the-scenes management is the invisible shield protecting our digital world. It encompasses robust, real-time monitoring systems that can detect anomalies and potential threats, strategic spares management for critical components, tight integration of subsea and terrestrial networks to ensure seamless data flow, and clearly defined long-term governance structures that mandate maintenance and upgrades. It is this comprehensive, often unseen, effort that ensures a single cable cut—whether caused by a rogue anchor, seismic activity, or even deliberate sabotage—does not escalate into a regional blackout or a catastrophic global disruption.
Crucially, this industry-led push for enhanced resilience is now being reinforced by progressive public policy. The recent adoption of the Enhanced ASEAN Guidelines for Strengthening Resilience and Repair of Submarine Cables represents a vital step forward. These guidelines, which notably recommend capping repair processing times at a maximum of 10 days, reflect a growing recognition among regional policymakers that rapid repair readiness is not merely an operational concern but a fundamental component of economic stability and national security. Such policy frameworks foster greater cooperation between nations, streamline regulatory approvals for repair operations, and encourage investments in repair capabilities, collectively strengthening the resilience of the entire network.
For businesses and policymakers alike, the practical consequence of this evolving landscape is a necessary change in what gets funded and prioritized. Compute capacity and data centers often take precedence in investment decisions because their tangible capacity and the immediate returns they offer are highly visible and easily quantifiable. In contrast, the intricate network of subsea cables, their associated landing stations, and the essential alternative routes that keep them connected are far less visible, and consequently, can be harder to justify as standalone investments. Yet, the trade-off is stark and clear: underinvestment in this foundational connectivity risks leaving expensive AI infrastructure vulnerable to crippling disruption, debilitating congestion, and significantly higher operating costs.
Therefore, connectivity must transcend its traditional categorization as merely an operational cost hidden deep within the IT budget. It must be recognized and treated as an essential and strategic asset, akin to national highways, power grids, or port infrastructure in the physical economy. The global digital economy, driven increasingly by AI, begins not in the cloud, but beneath the ocean. As AI adoption accelerates globally and within Asia, Singapore’s continued competitiveness as a premier global digital hub will fundamentally depend on the strength, resilience, and strategic investment in that invisible, yet utterly indispensable, lifeline beneath the sea. The future of digital prosperity hinges on acknowledging this profound truth and acting upon it with urgency and foresight.
The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.
