A massive outage at Amazon Web Services (AWS) disrupted thousands of websites and applications globally, including Snapchat and major banks, but has been resolved after hours of widespread impact, highlighting vulnerabilities in cloud infrastructure.
The outage began early on October 20, 2025, with user reports starting around 07:00 BST and peaking at over 11 million global incidents tracked by Downdetector. It affected more than 1,000 apps and sites, spanning social media platforms like Snapchat and Reddit, gaming services such as Fortnite and Roblox, and financial institutions including Lloyds Bank and Halifax in the UK. Many users found themselves unable to access critical services, from online banking to entertainment, causing frustration and operational halts across various sectors.
Amazon identified the root cause as an issue with DNS resolution for the DynamoDB API endpoint in its US-EAST-1 region, which acts as a core component for translating web addresses into machine-readable codes. This failure triggered cascading problems across AWS services, leading the company to throttle certain network functions to mitigate the disruption. Despite initial recovery efforts, the outage persisted for much of the day, with Amazon confirming full restoration only by late evening BST.
Real-world consequences were immediate and far-reaching. Airlines like Delta and United experienced flight delays due to system failures, while businesses reliant on AWS for internal operations saw significant productivity losses. Smart home devices, including Amazon’s Alexa-enabled products, became unresponsive, and streaming services such as Disney+ and productivity tools like Canva faced accessibility issues. User reports highlighted difficulties in completing everyday tasks, from mobile payments to accessing work materials.
Experts emphasized the concentrated dependency on cloud providers as a critical risk. Prof Alan Woodward of the University of Surrey noted the interconnected nature of modern infrastructure, where minor errors can propagate widely. Mike Chapple from Notre Dame University compared the outage to a power grid failure, suggesting that initial fixes may have addressed symptoms rather than underlying causes, leading to prolonged recovery times.
The financial impact is projected to be substantial, with estimates from Catchpoint CEO Mehdi Daoudi indicating potential losses in the hundreds of billions of dollars due to halted business activities and reduced productivity. This incident underscores the economic fragility tied to dominant cloud platforms, as AWS alone accounts for a significant portion of the global cloud market, serving major clients from Disney to the U.S. Army.
In response, voices like Cori Crider of the Future of Technology Institute called for diversifying away from monopoly providers to enhance resilience. She argued that over-reliance on a handful of companies, including Amazon, Microsoft, and Google, poses risks to security and economic stability, advocating for more localized and redundant systems to prevent future large-scale disruptions.
With services now restored, the outage serves as a stark reminder for organizations to implement robust backup plans and redundancy measures. As cloud computing remains integral to digital life, ensuring greater infrastructure resilience will be crucial to mitigating the impact of similar events in the future.