Seven major tech incidents in a week – what’s going on, and what can we do?

Seven major tech incidents in a week – what’s going on, and what can we do?

After a string of major tech outages and incidents in the past week, it’s hard not to ask: is this really just a coincidence?

Microsoft’s global outage last week disrupted email and business communications. A nationwide US aviation ground stop disrupted thousands of flights. Transaction problems at JPMorgan Chase affected the customers of other American banks. Apple suffered a services outage. Cyber criminals used legitimate third-party software Optimizely to plant malicious code on a slew of popular websites, including the BBC, CNN and Zoom.

All within days.

We know generative artificial intelligence (AI) has made sophisticated hacking tools increasingly available. So it makes sense to ask if these disparate incidents can be traced to AI-assisted cyber crime.

However, what we know so far points to a more complex picture, with no simple solution.

Despite it looking like the digital sky is falling, there’s no evidence the outages were related. But there does seem to be an underlying pattern of organisations patching and upgrading older systems.

Microsoft’s preliminary investigation of its outage points to a software defect that remained hidden for months. US aviation authorities linked the ground stop to newly installed network equipment.

Apple attributed its disruption to an internal update. On Wednesday, the US National Security Agency reported problems with its classified internal network, but said the outage happened during scheduled maintenance.


Author provided

The only cyberattack in the list is the Optimizely software being compromised. This online marketing platform, used by many major websites, was hijacked to steal customer payments and redirect about US$7,000 to cryptocurrency payment addresses, capture customer card details and exploit vulnerabilities in older iPhones.

This actually shows that cyber criminals don’t always want to bring systems down. While millions struggled with failing digital services in the past week, a cyberattack quietly unfolded through trusted software on more than 170 websites.

Why so many tech failures at once, then?

Imagine a self-driving car that performs perfectly in a simulator, only to encounter unpredictable traffic, pedestrians and roadworks in a crowded city. Real roads are more complicated than any test environment. And a hidden defect can affect a whole fleet of vehicles at once, causing chaos.

Modern digital systems – from aviation networks to banking to big tech suppliers like Microsoft – face the same challenge.

Tech systems need updates for various reasons, from security patches and fixing bugs to introducing new features. And these updates are tested “offline” before engineers deploy them.

However, an update to a system may pass every test, but behave differently across thousands of interconnected servers handling millions of requests. You might recall the massive Windows outage in mid-2024 that was traced back to a single software update.

Automation can magnify these problems. Computers can deploy updates, monitor servers and redirect traffic without waiting for humans to intervene. The automation that allows companies to manage thousands of servers in minutes can also spread one mistake across thousands of servers in minutes.

So how can we keep things running?

The ability to keep essential services working through a disruption, whatever its cause, is known as operational resilience.

Governments are beginning to act. Australia has sovereignty requirements for government data hosting (which means keeping some sensitive data onshore). In June, the government signed a digital resilience agreement with Microsoft, which means working with the tech giant more closely to address potential cyber risks as they come up.

But such measures can’t keep essential services running while a global provider is down.

Australia doesn’t need to build its own Microsoft or Google. Replicating these global platforms would be costly and complex, and wouldn’t guarantee freedom from outages.

Returning to pen and paper isn’t an option either. Paper records can keep a pharmacy counter running for a few hours, but essential services now handle volumes of data that no manual system could absorb for long.

Governments have a duty of care to keep essential services running. That means investing in what I call “targeted sovereign technological capabilities”, to allow select essential services to continue operating under disastrous conditions.

This means establishing a minimal suite of options we can run locally. These would be pre-configured, regularly tested local systems that can keep the core functions of hospitals, emergency services and other critical areas running when global platforms such as Microsoft fail or are attacked.

Regulators should require operators to test which critical functions they can carry out without their usual providers, and for how long, building on existing rules.

We can do more

Two further steps would help.

First, Australia should have a say over how global providers roll out changes to their essential services, not merely be told about them. The rule should be universal: no single change should reach all of a nation’s essential services at once, with urgent security fixes the exception.

Second, independent review shouldn’t stop at cyberattacks. Australia’s Cyber Incident Review Board examines significant cyber incidents, but almost every recent failure was declared not to be one, leaving the explanations to the companies themselves.

The post “Seven major tech incidents in a week – what’s going on, and what can we do?” by Mahmoud Elkhodr, Senior Lecturer in Information and Communication Technologies, CQUniversity Australia was published on 11/19/2026 by theconversation.com