Major AWS service outage disrupts Snapchat, ChatGPT, other platforms

Amazon Web Services experienced a major outage on the morning of October 20, disrupting multiple platforms that rely on its infrastructure, including Amazon, Alexa, Snapchat, Fortnite, ChatGPT, and more.

The company reported increased error rates and delays for several AWS services in the US-EAST-1 region.

User complaints surged after 7:30 a.m. London time, according to outage tracking site Down Detector.

AI firm Perplexity noted that the disruption affected the stability of its website.

No further details have been provided on the cause or resolution timeline.

🔗 Source: Bloomberg

🧠 Food for thought

Implications, context, and why it matters.

US-EAST-1 outages hint at design weak spots

The latest US-EAST-1 (AWS’s Northern Virginia region) disruption is at least the fourth since 2020, with incidents in 2021 and 2023 1. Elastic Compute Cloud (EC2) offers a service level agreement (SLA) promising 99.99% availability for deployments across Availability Zones (physically separate data centers within a region) 2.
Outages hit Alexa, Fortnite, Snapchat plus ChatGPT and the McDonald’s app 1. Perplexity named AWS as the cause 1, exposing limited redundancy for AI-native platforms built on distributed systems (software designed to run across multiple servers and locations).
Customers lack detail on whether it was a control plane (the management layer that provisions and orchestrates cloud resources) API failure or an infrastructure fault. AWS will issue a Post-Event Summary for significant control plane or infrastructure failures 3 to help teams judge credit claims under compute SLA thresholds 2.

Cloud management and FinOps (cloud financial operations) providers can ease SLA claims

AWS uses a tiered system that needs uptime logs and claims within two billing cycles 2 4. EC2 customers may qualify for 10–100% service credits based on downtime 2.
Multi-cloud migration consultants can spread workloads across regions and cloud providers, since 99.99% SLA commitments 2 did not stop hours-long disruption in US-EAST-1.
Monitoring and observability vendors can fill the gap from the 45-minute interval in AWS’s initial messages 1 with alerts to spot issues before AWS confirmation. Amazon CloudWatch (AWS’s monitoring service) carries a 99.9% SLA 5, so teams need independent monitoring to verify service credit eligibility.