The alarming reality of the internet blackout: As Microsoft Azure crashes just days after Amazon Web Services, experts warn we’re ‘seeing what putting all your eggs in one basket looks like on a global scale’

Hundreds of thousands of people were left unable to access critical services^[1] last night, as Microsoft^[2] Azure unexpectedly crashed.

The massive outage came just nine days after an Amazon Web Services outage brought ‘half the internet’ to a standstill.^[3][4]

Now, experts have revealed the alarming reality of the internet blackout. 

Microsoft and Amazon are the world’s two largest providers of ‘cloud computing’. 

The vast majority of internet services – including apps, social media platforms, and websites – rely on these companies to process their data.

Together, Amazon and Microsoft’s vast data centres, known as ‘hyperscalers’, host over 60 per cent of the world’s cloud services.^[5]

So when there is a problem, almost everyone around the world feels the impact.

‘We’re seeing what “putting all your eggs in one basket” looks like on a global scale,’ warned Colette Mason, AI consultant at Clever Clogs AI.

Hundreds of thousands of people were left unable to access critical services last night, as Microsoft Azure unexpectedly crashed. The massive outage came just nine days after an Amazon Web Services outage brought ‘half the internet’ to a standstill. Now, experts have revealed the alarming reality of the internet blackout

When services like Microsoft Azure or Amazon Web Services go down, they cause massive issues because of how companies structure ‘cloud computing’.

Cloud computing is essentially where companies or individuals rent out IT infrastructure from one of the tech giants.

Instead of paying the massive costs of building and operating their own servers, it’s usually cheaper for companies to pay for data to be processed for them.

However, this means massive parts of the internet rely on a few centralised ‘nodes’ where the world’s data is processed.

‘We have become reliant on the big three or four cloud providers because of cost and simplicity,’ Patrick Burgess, of BCS, The Chartered Institute for IT, told Daily Mail.

‘They make it affordable to provide new services at a scale that it would be difficult to replicate as individual organisations.’

Experts say that the total dominance of Amazon Web Services, Microsoft Azure, and Google Cloud is now creating serious risks.

Dr Saqib Kakvi, of Royal Holloway University, told Daily Mail: ‘The majority – around 60 per cent – of companies will be relying on a near triopoly of Amazon AWS, Microsoft Azure, or Google Cloud.

Microsoft Azure is the cloud computing service which provides servers and hosts data processing for thousands of companies, including Starbucks, Xbox, Kroger, and Costco 

Experts say these issues stem from the fact that the world’s cloud computing services are concentrated in the massive data centres operated by Amazon, Microsoft, and Google. Pictured: Amazon Web Services’ data centre in Ashburn, Virginia 

‘This means that we are putting all our eggs in one of three baskets. 

‘Although these companies are leaders in cloud technology, they are not infallible.

‘As we have seen twice in the past 10 days, if even a single provider has an issue in part of its infrastructure, then hundreds or even thousands of services become unavailable, affecting millions of customers worldwide.’

Yesterday, starting from around 15:30 GMT, an outage affecting Microsoft Azure triggered over 105,000 reports of disruption on Down Detector.

The problems affected Microsoft services like Microsoft 365, Outlook, Xbox Live, and Copilot.

However, since so many companies use Microsoft Azure’s cloud processing, the disruption rapidly spread to companies like Costco, Starbucks, and data tools like Blackbaud.

According to Microsoft, the issue arose due to what the company called ‘an inadvertent configuration change’ in Azure’s Front Door content delivery network.

Just nine days earlier, Amazon Web Services went down in an outage that affected more than 2,000 companies and millions of internet users around the world.

Microsoft’s outage comes just nine days after Amazon Web Services experienced another major outage that caused disruption for millions of users  

According to Amazon, the enormous disruption was triggered by a ‘malfunction’ at one of its data centres in Northern Virginia.

While giants like Amazon Web Services and Microsoft Azure are generally more reliable than smaller firms, experts say these outages highlight the problems of being over–reliant on just a few companies.

The results of just one firm going down were massive, but if multiple providers experienced issues at the same time, the results could be far worse.

Dr Kavi says: ‘Large numbers of websites and apps would be unusable, and this could include critical services such as banking, transport and logistics.

‘Dependencies between these could lead to downtimes of days rather than hours, as with the recent outages.’ 

In fact, these three providers are now so thoroughly embedded in the structure of the internet that people can be affected even if they don’t rely on cloud computing services.    

Professor James Davenport, a computer scientist from the University of Bath, explained to the Daily Mail: ‘Companies buy in services from other companies and, very often, those don’t disclose the source of the reliability they have.’

Since the internet is so reliant on a handful of individual firms, this domino effect means small failures have widespread, and often unexpected impacts.

Experts say that the massive dependence of Amazon and Microsoft means that small errors can cause major disruption to multiple services. During the Amazon Web Services outage, this included GOV.UK

Collectively, Amazon and Microsoft host around 60 per cent of the world’s cloud processing. This makes the global internet extremely vulnerable to blackouts when errors occur  

For example, the Amazon Web Service outage led to problems with electric locks, beds, and even ovens that relied on software connected to the cloud^[6].

Dr Jongkil Jeong, of the University of Melbourne, told Daily Mail that these outages ‘highlighted a fundamental vulnerability in how cloud computing is currently structured.’

‘This high reliance means that a problem at one vendor can paralyse a massive portion of the internet,’ he added. 

What makes this problem so difficult to resolve is that the massive power of these few companies means they are able to control the market.

While the tech giants secured their dominant position by offering companies an unbeatable deal, there are now concerns that these companies are stifling competition.

Dr Jeong points out that these services impose ‘vendor lock–in’ by making it prohibitively difficult, expensive, and complex to change providers.

For example, Amazon Web Services and Microsoft Azure often charge ‘data egress costs’ – fees for moving your data out of their platform to a new service.

This means that competitors find themselves locked out of the market and unable to compete with the bigger players.

Since Amazon Web Services is used by so many different services and pieces of software, even companies that don’t pay for cloud processing can be taken down during an outage 

To make matters worse, experts say that these firms’ near monopoly encourages them to ‘run hot’ – meaning they don’t use any more resources than they need to.

In July, the Competition and Markets Authority (CMA) said its investigations into the UK cloud services market showed it was ‘not working well’.

The regulator recommended using its powers to designate Amazon and Microsoft as having ‘strategic market status’ – allowing it to demand changes to boost competition.

Nicky Stewart, Senior Advisor to the Open Cloud Coalition, told Daily Mail: ‘This dependence stifles competition and innovation while creating a fragile system where one provider’s issue can quickly become everyone’s problem.

‘The market is highly anti–competitive and very difficult for challenger cloud providers to even have a voice, let alone get an opportunity, leaving companies with few realistic alternatives.

‘Europe’s cloud sector urgently needs bold regulatory action to break this dependence and restore real competition.’ 

An Amazon Web Services spokesperson told Daily Mail: ‘For nearly 20 years, AWS has maintained a track record for reliability that surpasses other major cloud providers, and customers choose AWS because we provide the most innovative, secure, and scalable services.’

Microsoft declined to provide a comment in response.