Amid the drive to scale artificial intelligence (AI), modernise legacy systems and deliver richer digital experiences, businesses are increasingly fragile if and when cloud services fail. The effects are often immediate: applications stall, websites slow or crash, and the knock-on financial impact can be severe.

When cloud and AI workloads become central to daily operations, even brief disruptions can create significant challenges. IDC’s Enterprise Horizons 2025 InfoBrief found that more than half of Asia-Pacific (Apac) organisations reported revenue losses exceeding US$5 million (RM19.5 million) in the past year due to outages or poor network performance.

 This is particularly pertinent for Malaysia, where digital dependence is near-universal. Internet penetration sits at around 98%, with nearly 97% of households online. Cloud outages here have the potential to affect almost the entire population.

An online-first population has catapulted Malaysia’s digital economy into a major engine of growth, accounting for 23.4% of gross domestic product (GDP), RM451.3 billion in 2024 alone. In just the first nine months of 2025, e-commerce revenue had already reached RM937.5 billion, a 1.9% rise from the same period in 2024.

With more services moving online, expectations for speed and reliability are higher than ever. By July 2025, 5G reached 82.4% of the country, supported by 7,489 transmitter sites, while internet access in populated areas hit 98.8%. Digital services have truly become part of Malaysia’s everyday infrastructure.

Recognising the stakes, Malaysia launched its first National Cloud Computing Policy (NCCP) in August 2025. The policy aims for a Whole-of-Nation approach, and businesses must do their part to create a secure, sovereign, and sustainable cloud ecosystem.

Cloud computing brings many benefits, but depending entirely on a single provider is a growing vulnerability. Every major outage highlights just how risky a one-cloud approach can be. While keeping everything on one platform may seem convenient at first, over time, it creates a single point of failure.

The issue is not cloud adoption itself. Rather, it is how deeply organisations depend on it without planning for interruptions. Resilience, therefore, must be treated as a first-order design principle.

For many teams, multi-cloud is a natural response. In practice, this means distributing applications and workloads across multiple providers to reduce concentration risk.

It lets organisations run workloads where they work best, whether that’s for speed, staying compliant, keeping costs in check, or ensuring they are always available. This approach is especially useful as AI workloads grow, each with its own unique demands for computing power and low latency.

That said, multi-cloud isn’t without its challenges. Its main strength lies in spreading workloads across different providers, but that also adds complexity. Traffic needs to be routed smartly to keep services fast and available, and data has to stay consistent across different systems, even though each provider handles storage, networking, and resilience in its own way.

Every cloud works a little differently, with its own patterns and limitations. Planning for these differences from the start is essential, and ignoring them can create fragile systems that break exactly when you need them most.

When multi-cloud is set up correctly, it can make a world of difference. Traffic can shift smoothly between regions or providers during an outage, so services stay online even when parts of the system stumble. 

Redundant storage and cross-cloud backups keep data safe, while network protections, like distributed denial of service (DDoS) mitigation, help shield services from unexpected failures.

Organisations can avoid revenue loss, maintain operational continuity, and meet regulatory expectations. Perhaps most importantly, they gain the confidence to operate in a world where perfect uptime is never guaranteed.

As digital operations and AI workloads become more interconnected, resilience has to be a strategic priority. Experienced teams treat multi-cloud as a commitment to clarity, visibility, and thoughtful architecture. They design for imperfect conditions, plan for failures, performance fluctuations, and provider-specific constraints.

In Malaysia, reliable digital services underpin economic continuity and public trust. Decisions made by enterprises ripple across the broader digital ecosystem. Organisations that invest in robust, multi-cloud-ready architectures are better placed to navigate uncertainty while safeguarding revenue, customers, and operational stability.

 Rachel Ler is area vice-president of Asia at Fastly.