A widespread outage at Amazon Web Services (AWS) on Monday morning temporarily crippled major parts of the global internet, taking down websites, banking platforms, smart devices, and even some government services. Amazon said the issue, which began around 3 a.m. Eastern Time, has now been “fully mitigated,” with most systems returning to normal operations.

DNS failure: The outage was traced to problems with the Domain Name System (DNS). The failure meant that while data stored on AWS remained safe, millions of websites and applications could not locate their servers and went offline for several hours. Experts said DNS failures are particularly disruptive because they affect access to multiple dependent systems and can take longer to stabilize.

Widespread impact: Among the affected were major online platforms such as Coinbase, Fortnite, Signal, Zoom, and Amazon’s own Ring smart-home and surveillance devices. Airlines including United and Delta, several UK banks, and companies relying on AWS infrastructure also faced downtime. Downdetector recorded over 6.5 million user reports globally, with the United States and United Kingdom hardest hit.

AWS engineers worked through the morning to restore operations, gradually bringing services back online and advising users to clear cached data to speed up recovery. While some minor throttling and latency persisted, Amazon confirmed the “underlying DNS issue” was resolved and services had stabilized.

Industry experts warned that the outage highlights the Internet’s growing fragility and dependence on a handful of cloud providers. With AWS controlling roughly 30% of the global cloud market, a single regional fault can produce cascading effects across countless apps and organizations. “We’ve built an ecosystem where one provider’s failure ripples through the global economy,” said Professor Feng Li of Bayes Business School.

Future lessons: The company has yet to release a detailed explanation of what triggered the DNS malfunction, though it has ruled out a cyberattack. The incident follows previous large-scale internet disruptions, including the 2024 CrowdStrike software glitch that paralyzed systems worldwide and a 2021 Akamai DNS failure that briefly knocked major websites offline.

Monday’s AWS outage, though relatively short-lived, serves as another reminder of the concentration risks in today’s cloud-driven digital infrastructure—and how easily a single technical fault can bring much of the connected world to a halt.

Update – October 22: On Monday afternoon, Amazon declared the incident resolved. Engineers had traced the disruption to a DNS malfunction in its US-EAST-1 region—the firm’s oldest and most heavily trafficked data centre cluster. Though the issue was swiftly addressed, its ripple effects laid bare the fragility of digital infrastructure underpinning modern commerce. Analysts now liken the event to July 2024’s CrowdStrike debacle, with estimated damages running into the billions. The outage has reignited calls for “multi-cloud” resilience strategies, particularly among risk-averse industries like finance. For AWS, which powers swathes of the Internet from streaming video to airline bookings, the recovery may prove more reputational than technical.