Today’s global economy is deeply dependent on cloud services – the EU estimates that nearly 53% of its businesses use paid cloud services of some kind, with some countries reaching nearly 80%. Almost across the board, this dependency has grown rapidly even within the last few years.

Cloud services underpin a host of vital operations – DNS resolvers hosted by AWS allow users to access websites, Email servers that run through Google are vital for communicating with customers and colleagues alike, and data storage in services like SharePoint is necessary for holding vital client files.

The more functions that businesses upload to the cloud, the more disruptive any disruption can become.

Although each business uses cloud differently, suddenly having no access to cloud services is almost always disruptive, chaotic, and sometimes paralyzing for businesses. Unfortunately, this is hardly a remote possibility, and large-scale outages are becoming fairly common – with some declaring that they have become a ‘new normal’ in 2026.

Parametrix’s cloud outages risk report 2025 found that there had been 45 critical interruptions of cloud service lasting over 150 hours in total. Even outages that last only a few hours can do massive cumulative damage to the economy.

A recent AWS outage in the Autumn of 2025 that lasted 15 hours led to estimated losses between $38 - $581 million in material damages. When factoring in long-term damage to reputation, revenue and productivity, indirect damage from cloud outages can reach into the billions a figure that only seems to be climbing, even as major data centers invest in their security.

Companies that have not invested in the preparedness that would allow their business to continue when they don't have access to key cloud functionality are finding their business grinding to a halt, losing customers, reputation and tanking employee morale.

Why are these events becoming more damaging?

Aside from the fact that businesses are increasingly dependent on the cloud, outages are becoming more threatening because of market, infrastructure, cybersecurity and political factors. These factors are converging not only to increase the risk of outages themselves, but to increase the severity and damage to business when they do.

Market

Cloud computing is dominated by three major US ‘hyperscalers’ - AWS, Microsoft, and Google, who together own over two-thirds of the global cloud market. At present, there seems to be very few companies able to challenge the major players within the cloud market given the sector's reputation for being incredibly capital intensive.

This market structure effectively creates chokepoints in the world’s ability to access the cloud. The scale of these companies means that small internal errors within company systems can cascade into massive, market wide disruptions in cloud access.

Infrastructure

Concentration risk is only one part of the problem. The physical infrastructure supporting cloud growth is also under increasing strain.

While it's easy to think of “the cloud” as ethereal, it actually requires a complex web of interdependent physical infrastructure – especially power, which data centers need vast amounts of. In areas where sufficient power is unavailable, or the grid has become over-taxed, critical failures are a much more present threat.

Power outages are the second most common cause of critical cloud outages.

This is equally true of water access – data centers need constant access to fresh water to cool their systems. As climate change, community resistance and infrastructure limits these resources, it introduces the risk of catastrophic total failure.

Security

Data centers are becoming targets during political conflicts – most recently the war in Iran, where several data centers were targets. States are beginning to realize the importance of data centers and the leverage they represent.

It’s a key reason many states, including the EU, have begun data sovereignty initiatives that would build more cloud capacity within their borders. If states can cut off, restrict or freeze access to the cloud after a political dispute, all businesses become more vulnerable.

Data centers are incredibly appealing targets for criminal and national level hackers given the amount of critical data and operations stored there. The UK's National Cybersecurity Centre has emphasized the importance of building more resilient data center software given increasing numbers of attacks.

It’s unavoidable that as businesses, governments and militaries become more dependent on the cloud, data centers will become targets of bad actors.

How businesses can de-risk themselves

Cloud outages can be particularly devastating for businesses and need to be treated like an ongoing strategic risk, not just a one-off disaster scenario. While businesses can never eliminate the risk of an outage, there’s a lot they can do to reduce its impact.

Given that migrating completely off the cloud is not an option, businesses need to start actively planning what outages and cloud downtime might do to their business and plan around it, a form of Business Continuity Planning (BCP).

Businesses need a clear conception of how their business will continue to operate if they have limited (or zero) access to the cloud.

Some examples include:

-Using local backups or edge computing that would store key data on-site, or to create larger amounts of workflows that can be done completely offline.

-Distributing your systems across multiple regions within a cloud provider – given that many outages are regional specific; this will shield a business from common incidences and requires the least amount of setup. However, it doesn’t shield the significant problem of being too dependent on the performance of a single cloud provider.

-Using multiple different cloud providers helps reduce this exposure. By being operable on multiple different cloud providers, it is insulated from operator specific outages while maintaining the benefit of cloud connectivity.

To do this effectively, however, a company will need to maintain some synchronization between their solutions to keep business continuity, either manually or by using third-party software.

Looking forward

Cloud outages are now a fact of life, and businesses that depend on cloud access need to build effective adaptation plans to avoid significant disruption. As businesses continue to rely on the cloud, they can no longer afford to treat their resilience plans as a ‘nice-to-have' option.

We've featured the best cloud database.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit