Amazon Web Services Outage Cripples Global Internet, Disrupts Millions of Users

Amazon Web Services’ US-EAST-1 (Northern Virginia) region experienced a major outage on October 20, 2025.

“By:Express Web Desk Amazon said on Monday evening that a cloud computing unit at its data center in northern Virginia has largely resolveda massive outage that disrupted the internet usage worldwideand caused thousands of websites, including popular apps like Snapchat and Reddit, to face global outage,Reutersreported”

The Indian Express

This disruption affected parts of the global internet for several hours.

The Indian Express

Multiple sources identified the cause as a DNS-related failure linked to AWS’s internal networking.

Reports indicate that a subsystem responsible for monitoring network load balancers within EC2 malfunctioned.

This malfunction caused a spike in error rates and latency, which blocked access to DynamoDB’s API.

AWS began addressing the issue within hours and declared services fully restored by early evening Eastern Time.

However, the exact timelines of recovery vary slightly across different reports.

The incident highlighted how a fault in a single region, especially AWS’s oldest and largest cluster, can have widespread effects across the internet.

The disruption rippled through thousands of apps, websites, and essential services across sectors including banking, airlines, media, gaming, and government.

Millions of user problem reports were logged worldwide as a result.

Houston Public Media

Popular platforms such as Snapchat, Roblox, Venmo, Coinbase, and Amazon’s Prime Video and Alexa experienced difficulties.

UK institutions including Lloyds, Halifax, Bank of Scotland, and HMRC also reported issues.

Some reports highlighted that even support channels within AWS were impaired, complicating incident response for customers.

Accounts diverged on the precise root cause, though they broadly converged on DNS/DynamoDB interactions and internal network health monitoring.

“A major global outage on Monday, October 20, 2025, disrupted access to numerous popular websites and services, including Snapchat, Fortnite, Duolingo, and several UK banking apps like Lloyds and Halifax”

Daily Mail

Several outlets—especially local and public media—reported a faulty software update to DynamoDB in Northern Virginia as the trigger.

Others focused on DNS resolution failures and EC2 load balancer health checks.

Meanwhile, most mainstream and security sources rejected cyberattack theories.

However, a few tabloids expressed uncertainty about the cause.

Even after AWS said the core issue was mitigated, recovery was uneven as backlogs cleared and some services kept failing.

Several reports note throttling of new EC2 instance launches and lingering errors in Lambda and other services.

Deadline

Others describe message queues and configuration services needing hours to catch up.

That gap between ‘fixed’ and ‘fully recovered’ prolonged user pain and operational delays across industries.

The outage renewed warnings about the fragility of centralized cloud dependence and spurred calls for redundancy, multi-cloud strategies, and stronger oversight.

“On October 20, beginning around 8 a”

Cambridge News

West Asian and regional outlets focused on practical resilience by advocating multi-cloud and failover solutions.

Cambridge News

UK coverage pressed regulatory questions about designating AWS as a Critical Third Party and flagged potential liability disputes.

Sector-specific reporting emphasized that industries like cryptocurrency and finance face outsized risks if a single provider experiences failure.

Consumer guidance urged households to prepare basic contingencies in case of similar outages.

Early cost estimates and market reactions diverged significantly.

Some outlets warned of staggering economic damage from lost productivity and halted operations.

Others noted investor resilience and even a slight uptick in Amazon shares.

Analysts compared the event to last year’s CrowdStrike meltdown, framing it as another wake-up call on systemic concentration risk in cloud infrastructure.