Skip to content
REALNEWS HUB

Why a week of major tech outages hit Microsoft, US flights and banks

There is no evidence these are linked, but the constant series of tech disruptions can be unnerving.

REALNEWS HUB Newsroom

Server rack corridor in a data centre illuminated with blue LED lights, showing glowing cables and hardware units.
REAL NEWS HUB

A run of large technology failures in a single week has led many people to ask whether they are connected. BBC News reports that there is no evidence of a link, but that the incidents together show where the modern tech world is vulnerable.

In the past week there was a major Microsoft outage, an apparent glitch that kept every US flight on the ground for hours, and trouble at the National Security Agency. Some of the biggest US banks also suffered disruption, and a serious flaw turned up in a widely used programming language.

According to the BBC, many of the failures began with an update or installation that went wrong. A network upgrade at Microsoft overloaded its servers and caused them to shut down. Upgrades to an ageing network were behind the grounding of thousands of US aircraft at the weekend, and the BBC says engineers were in the middle of swapping old copper cabling for fibre. The NSA's internal communications problem, and separate faults with Apple's iCloud and Apple Pay, were likewise put down to updates. The payment problems at US banks were blamed on a "software bug" at JP Morgan Chase, which says it was not a cyber-security incident.

Other problems do involve security weaknesses. AI was used to uncover flaws in Java, the third most popular programming language in the world. Researcher Scott Helme also found malicious code in Optimizely, a platform that some of the biggest companies rely on. The BBC says hackers took roughly $7,000 (£5,300) in cryptocurrency before they were detected and expelled.

Speaking on BBC Radio 4's Today programme, Helme said he would be "very surprised" if attackers were not using AI to speed up their work. The BBC notes that since the summer, AI companies including Anthropic and OpenAI have held back their most advanced models from the public, giving selected companies and institutions access first. That lets security professionals find flaws and alert others before criminals get there. The BBC says this is what happened with Java and with a separate PHP flaw, where there was a race to fix the problems before they could be exploited. For PHP, a single volunteer programmer worked to produce patches that were simple to apply.

The BBC points out that outages are not new, citing earlier worldwide disruption from CrowdStrike and Amazon Web Services. It suggests technical debt may play a part: quick fixes used in place of dealing with an underlying problem can pile up and cause bigger trouble later. Companies rethinking their cyber-security because of AI may also be pushing out more patches, and every update brings a risk of disruption.

US Secretary of State Marco Rubio has dismissed the idea of a coordinated attack. "We are not under systemic attack," he said, according to the BBC, adding that the events are not inter-related. Speaking to reporters in Manila on Wednesday, he said there was "zero evidence" of a connection, CBS News reports, and that the administration has no reason to believe hackers, criminals or a foreign state were behind any of the incidents.

The impact reaches well beyond the screen. The BBC says outages frequently disrupt healthcare, schools and people's ability to do their jobs, and at the weekend they stopped planes flying. They also highlight how much the world leans on a handful of companies. Microsoft, Amazon and Google run a large share of the world's servers, and Oracle's Java is used by almost all of the largest firms.

Sheila Pancholi, risk assurance lead at RSM UK, described the Java flaw as a reminder of how reliant businesses have become on "a relatively small number of critical technologies". She said organisations should know where their critical technology dependencies are, respond fast when weaknesses appear, and check that the response worked. Xander Van Raemdonck, a security engineer at Aikido Security, said the difficulty lies in "knowing whether you're exposed". Java is built into a great deal of business software, he said, so many companies have no complete list of where it is running.

The BBC says this may be the new normal for now. The AI tools are new enough that they will probably keep finding flaws that have sat in code for decades. Martin Reynolds, field CTO at Harness, said AI is speeding up discovery for both attackers and defenders as well as the amount of software being released, so developers face more to patch with less time. Helme said attackers can move fast and experiment without fear of failure, and that a small but visible gap currently separates their speed from that of defenders.

Earlier on this story

  1. Java flaw found by AI prompts urgent patch calls in US and South Korea
  2. Rubio says no evidence links US tech outages, rejects claim of systemic attack
  3. NSA internal network outage disrupts staff communications amid wave of US tech failures
  4. Senators urge FAA to halt router replacements after nationwide ground stop
  5. United Airlines hit by system outage affecting departures days after US ground stop
Full coverage: Cybersecurity & Data Breaches
Explained: catch up on this storyHow did hackers steal $6.3bn from Binance?Attackers compromised Binance using deceptive recruitment interviews and an internal artificial intelligence system, draining $6.3 billion in cryptocurrency in an assault US authorities link to North Korea.Read the explainer →

Spotted an error? Report it · Editorial standards · Corrections policy

Source & verification

Developing

Reported from BBC News.

How we verify and rate stories

Get breaking news alerts and the Morning Briefing

Verified stories in your inbox, every one linked to its sources. All newsletters →

Newsletters

Free. Unsubscribe any time. See our privacy policy.