On Monday, October 20 2025, a big internet outage emanating from the Amazon Web Services (AWS) US-East-1 region in Northern Virginia (USA) caused widespread disruption to major apps and websites worldwide.
AWS first reported increased error-rates and latency at 3:11 a.m. US Eastern Time (ET) (07:11 UTC) in the US-East-1 region. By around 8:00 a.m. British Summer Time (BST) (07:00 UTC), users across Europe and elsewhere were already reporting service failures. Within hours most services had resumed, with AWS stating “significant signs of recovery” by 11:00 a.m. BST (10:00 UTC).
Scope and scale of the disruption
The outage impacted a wide array of platforms that rely on AWS cloud infrastructure:
- The social-media app Snapchat, gaming platform Fortnite, and the retail site Amazon were among the most-visible services affected.
- Other affected services included Signal, Duolingo, as well as financial platforms like Venmo, Robinhood and AI-startup Perplexity.
- In the UK, major institutions were affected including Lloyds Bank, Bank of Scotland and HM Revenue & Customs (HMRC).
According to the outage-tracking service Downdetector, thousands of reports flooded in starting around 7:30 a.m. London time (02:30 ET) on Oct 20.
Cause and cloud-infrastructure context
AWS’s initial statement said:
“We can confirm increased error rates and latencies for multiple AWS Services in the US-EAST-1 Region.”
Further reporting identified the root issue as a DNS-resolution disruption of the DynamoDB endpoint in US-East-1, which triggered failures across other AWS services including CloudFront, Step Functions and Security Token Service.
Because many apps and websites rely on AWS’s US-East-1 region—even when operating globally—the failure in that region had a cascading effect worldwide.
Impact for users and businesses
For users, this big internet outage meant:
- Inability to log into apps like Snapchat, Fortnite and Signal or perform key tasks.
- E-commerce disruption: Amazon’s checkout and other retail services slowed or failed.
- Banking and government services in the UK and elsewhere were temporarily inaccessible.
For businesses:
- Lost transactions and customer frustration due to service downtime.
- Highlighted the vulnerability of depending heavily on a single cloud region/provider.
- Raised urgent questions about resilience, redundancy and multi-region strategies.
Cybersecurity and cloud-infrastructure experts noted that today’s incident reaffirms that cloud-based failure is a systemic risk when too much depends on one region.
Recovery and what comes next
By 10:27 BST (09:27 UTC), AWS reported “significant signs of recovery” and said most requests should now be succeeding. The US-East-1 region continues to be monitored while queued requests are cleared. Some secondary services remain impaired but the worst is past.
Going forward:
- Businesses may review cloud-deployment strategies and consider multi-region or multi-cloud setups.
- Users may experience lingering effects (slower login, partial outages) for parts of the day.
- Industry watchers will test how cloud-providers respond and what transparency improves.
FAQs
The outage began on Monday, October 20 2025 around 3:11 a.m. US Eastern Time (07:11 UTC) in the AWS US-East-1 region (Northern Virginia, USA).
Major platforms hit include Snapchat, Fortnite, Amazon (retail/voice-assistant), Signal, Duolingo, Venmo, Robinhood and others. UK entities such as Lloyds Bank, Bank of Scotland and HMRC also faced issues.
Yes — by around 10:27 BST (09:27 UTC), AWS said most affected services were recovering and many are already back online. Residual issues may persist for some services.
