Amazon Web Services experienced two significant outages in July 2026, disrupting critical internet services and highlighting concentrated infrastructure risks within cloud computing. The incidents—CloudFront outage on July 16 and a major regional failure on July 24—demonstrated how single points of failure in AWS infrastructure can cascade across millions of dependent services.
The first incident involved AWS CloudFront, Amazon’s content delivery network that caches and serves web content globally. On July 16, CloudFront customers using VPC Origins experienced a 3.5-hour outage affecting websites and services relying on the platform. Amazon identified an internal connection-management limit as the root cause, preventing routing configurations from loading correctly and disrupting VPC Origin connections. High-profile services affected included HuggingFace, the UK National Lottery, Tailscale, and Ubiquiti.
The July 24 outage proved far more impactful. Beginning around 7:40 AM ET Friday, AWS’s US-WEST-2 region in Oregon experienced regional internet connectivity problems that cascaded across seven major services. Initially AWS reported five affected services before expanding the list to include Global Accelerator and Elastic Container Service. The outage lasted approximately 80 minutes from first reports to AWS’s resolution notice.
Millions of users discovered services were unreachable as major platforms went offline simultaneously. DoorDash ordering, Reddit browsing, Apple Pay transactions, Hulu streaming, PlayStation Network gaming, and numerous other consumer-facing services stopped responding. The common thread connecting these seemingly unrelated outages was their dependence on AWS’s US-WEST-2 region. When that region experienced internet connectivity problems, the cascade affected every dependent service. PlayStation Network’s vulnerability to third-party cloud outages has been a recurring issue for gaming platforms worldwide.
This outage pattern reveals fundamental concentration risks in internet infrastructure. AWS holds 40%+ of the cloud market share. This dominance means single AWS outages impact far more internet services than comparable outages at competitors. When AWS goes down, it’s not just Amazon customers affected—it’s millions of services and billions of users worldwide who depend indirectly on AWS infrastructure.
The incidents marked AWS’s third notable reliability problem within three months. May 2026 brought a data center thermal event. June saw a network disruption. July brought two outages of varying severity. This cluster of incidents within a short timeframe raised questions about AWS infrastructure resilience despite Amazon’s massive investments in redundancy and reliability. The pattern is not unique to AWS: Meta’s 12,000-user outage earlier this year similarly exposed how centralized platforms amplify outage impact.
AWS maintains data centers across multiple geographic regions specifically to prevent failures in one region from affecting others. Yet the regional outage cascaded anyway, affecting numerous services simultaneously. This suggests AWS’s redundancy approach, while sophisticated, proves insufficient for certain failure modes. When regional internet connectivity fails, services using only that region experience simultaneous unavailability.
The outages renewed industry discussions about multi-cloud strategies and reducing dependence on single providers. Companies operating solely on AWS infrastructure faced complete service disruptions. Those using multiple cloud providers or hybrid approaches continued operating during AWS incidents. However, the practical reality remains that migrating away from AWS requires substantial effort and expense, creating vendor lock-in despite reliability concerns.
AWS typically publishes detailed postmortem analyses after major incidents, explaining failure causes and remediation steps taken. These transparency reports help customers understand what happened and inform business continuity planning. However, the incidents underscore that even massive technology companies with sophisticated engineering teams experience disruptions.
For internet users and organizations, the lesson is clear: cloud provider reliance carries concentration risk. AWS’s dominance means outages have outsized impact. Diversification and backup planning become essential for critical services despite the added complexity and cost of maintaining multiple cloud providers.