Amazon Web Services are down globally: Takes down Perplexity, Snapchat, Alexa and more; here’s what the company has to say

Amazon

A large-scale outage at Amazon Web Services (AWS) on October 20 triggered cascading disruptions across the internet, taking down or degrading dozens of widely used services globally. The incident, traced to the US-East-1 region — the most critical AWS zone — caused failures in both Amazon-owned products and third-party applications that depend on AWS for compute, storage, identity or backend delivery. As a result, users worldwide experienced login errors, stalled apps, broken streaming sessions, delayed responses, and unavailable cloud features across multiple platforms.What is impacted

Because AWS powers both Amazon-owned products and external services, the outage propagated through a long list of platforms. On Downdetector, sharp spikes were visible for services including:
• Perplexity AI
• Snapchat
• Amazon Alexa
• Ring Security
• Roblox
• Amazon Prime Video
• Fortnite / Epic Games Store
• McDonald’s app
• Epic Games infrastructure
• Venmo
• Chime
• Blink
• Spectrum
• T-Mobile
• Signal
• Reddit
• Steam
• Coinbase

Many of these failures are indirect — caused by authentication, storage, or API dependencies on AWS. In enterprise contexts, login systems, dashboards, and database-backed apps relying on DynamoDB and IAM also saw degraded performance or request failures.

What Amazon has said

AWS’ official status page confirms “increased error rates and latencies for multiple AWS Services in US-East-1”. Engineers say they have identified a DNS-related issue affecting the DynamoDB API endpoint which cascaded across services built on that region. Mitigations are in progress and AWS reports “early signs of recovery”, though some retries may continue to fail until backlogs clear.

AWS has recommended customers retry failed requests and says parallel recovery paths are underway. Additional updates will be shared as restoration progresses.

Until full recovery, any platform tethered to AWS infrastructure — either directly or through third-party integrations — may continue to experience instability.