When Amazon Web Services, aka AWS and Microsoft Azure – the twin pillars of global cloud computing – faltered within a week of each other recently, the internet itself seemed to blink. From smart mattresses to social media chat,  banking apps to business critical tools, millions of people were suddenly reminded that “the Cloud” really is just ‘someone else’s computer’. And that computer can stop working when you need it most. 

For all its strengths, the cloud is after all just infrastructure – and a surprisingly centralised infrastructure at that. When it goes down, so does much of modern life.

What brought down the cloud?

On 20 October 2025, 6:49 AM (UTC), AWS’s US-East-1 region suffered a cascading failure that spread worldwide in minutes. It began with a bug in its DNS management system for DynamoDB – a technically “small” fault that triggered a massive outage across dependent services. Apps like Snapchat, Venmo, Alexa, Signal and Ring went offline; even enterprise platforms such as Slack and Salesforce saw disruption. Hours later, AWS engineers restored service, but the ripple effects lasted days for some users.

Barely a week later, Microsoft Azure faced its own crisis. A configuration change in Azure Front Door – Microsoft’s global application delivery network – caused widespread issues with Office 365, Teams, Xbox Live, and Azure-hosted applications. Microsoft quickly rolled back the update, but the episodes’ proximity to each other underscored just how fragile even the world’s largest platforms can be.

When Convenience Becomes Dependence

These outages were wake up calls. For most people, they were the moment the internet’s invisible plumbing became very, very visible. The services we use to shop, chat, work, and even control our homes all rely on a handful of hyperscalers – AWS, Azure, and Google Cloud chief among them.

The usual argument is that hyperscalers are the efficient way to host things, but they’re also dangerous – with a single point of failure now carrying global consequences. Yes, AWS and Azure have immense market share and unparalleled technical sophistication, but that concentration of power is exactly why the world feels it so keenly when they stumble.

When your marketing team can’t access Canva, it’s a headache. When a hospital, airport, or even a national defence system is brought down by the same outage, it’s a crisis. 

The Case for Decentralisation

These events reinforce a principle that technologists have long championed: the need for decentralised systems. During the outages, for example, Signal Messenger, which relies on AWS, went dark. In contrast, Element Messenger, built on a federated, self-hosted model, stayed online. Its users – from journalists to governments – could continue communicating securely and uninterrupted. Decentralisation worked exactly as designed: resilience through independence.

For businesses, this lesson applies far beyond cloud infrastructure. Building in redundancy, spreading risk across providers, and exploring distributed technologies are operational strategies which are about to receive a whole lot more attention. It’s the same philosophy that advises having your wifi and phone SIM from different providers. Safety through redundancy. Any expert providers of on-premise services will be rubbing their hands with glee.

Small Is Resilient — In Technology and PR

Beyond the tech sphere the same logic applies to how companies communicate. Just as smaller, federated networks weather storms better than monolithic systems, smaller, specialist PR agencies can navigate disruption better than conglomerates weighed down by bureaucracy and uniformity.

At Positive, we believe smaller is safer, stronger and smarter. We’re not dependent on a single monolithic way of doing things. We adapt, we move fast, have back-up options to our back-up options, and give our clients something the big players can’t: resilience, focus, and personal attention. In technology or in communications, decentralisation is a philosophy that makes the difference between staying on or doing dark.

What to do when the lights go out

Cloud computing will remain the backbone of digital life, but as these outages proved, it’s not infallible. The future will belong to those who can blend the scale of global infrastructure with the flexibility of decentralised design, and in the short term, we’re bound to see a spike in sales of failover tech, and for cloud providers who can offer multi-cloud offerings to provide just that. Whether in code, communication, or collaboration, diversity of operations isn’t just healthy. It’s survival.