Most people notice digital infrastructure only when it fails. In my years working with data centers, digital infrastructure and disaster recovery, one lesson has stayed with me: resilience is not defined by a plan or a high-performance facility. It is demonstrated by what continues to operate when a connection is lost.

That question became more tangible in September 2025, when Microsoft warned that some Azure users in the Middle East could experience increased latency after multiple undersea fiber-optic cable cuts in the Red Sea. The incident did not result in a region-wide internet outage, but it showed how services relied upon by people and organizations can be affected when international routes are disrupted.

For Saudi Arabia, the challenge is not to replicate the global internet as a closed national network. It is to determine which essential services must continue operating if international connectivity is degraded, and ensure they can do so.

The first step is to understand what must keep working. Government, financial, healthcare and education services, along with food systems and logistics, support daily life and economic activity. Not every application requires the same level of protection. Agencies and operators should identify their most critical services, map dependencies on networks, cloud platforms, data and suppliers, and establish recovery priorities before an outage occurs.

The next step is to build resilience into the routes and systems on which those services depend. This requires diverse international connections, alternative routing, and local hosting for applications and data that must remain available. Saudi Arabia’s Internet Exchange, SAIX, helps networks exchange domestic traffic within the Kingdom. It is an important foundation, but local traffic exchange alone cannot keep a service running if its application or a critical dependency remains overseas.

The same principle applies to data centers. Essential workloads require geographically distributed facilities, independent network paths, backup power and cooling systems, and regularly tested recovery plans. Protection should be proportionate to the importance of each service. Some critical workloads may require fault-tolerant facilities such as Tier IV data centers. However, a Tier classification describes data center infrastructure; it does not, by itself, guarantee resilient connectivity or service continuity. Those depend on the broader architecture and how it is operated.

Data must also be ready when it is needed. Organizations should know where essential data originates, who is responsible for it, how it is classified and who can access it. During an emergency, services should be able to access the minimum data and systems they need through secure local arrangements. The National Data Management Office’s frameworks provide a useful reference for common standards and governance, while each entity remains accountable for the data under its care.

Finally, readiness must be tested. Plans on paper cannot demonstrate whether users can sign in, whether applications can reach the data they require, or whether traffic can shift successfully to alternative routes. Relevant entities should conduct regular, coordinated exercises in controlled environments, simulating the loss of selected external routes or services. They should test backup systems, data recovery processes, communications and decision-making procedures, then address any weaknesses identified.

Saudi Arabia already has strong foundations on which to build. The Communications, Space and Technology Commission’s Saudi Internet Report 2025 found that 56 percent of internet traffic in the Kingdom was local. This is a useful indicator of domestic traffic localization, but it does not reveal how many essential services could continue operating without international connectivity. Building on this foundation will require continued investment in national networks, local hosting, data centers, cybersecurity and robust data governance.

Digital resilience is not achieved by withdrawing from the world or creating a duplicate global internet behind national borders. It comes from understanding which services matter most, reducing single points of failure, and demonstrating through regular testing that essential operations can continue under adverse conditions.

The true measure of resilience is not the number of backup systems shown on a diagram. It is what people can still rely on when an international connection fails. If a major route were disrupted tomorrow, which essential services would continue to operate, and have we tested that assumption in practice?

• Dr. Rafiq M. Jamalaldeen is an executive adviser in data and AI, digital transformation, data centers and digital infrastructure, crisis and risk management.