Amazon Web Services (AWS) suffered a major outage on Monday that disrupted access to some of the world’s most widely used digital platforms, including Amazon.com, Alexa, Snapchat, Fortnite, Coinbase, and Canva. The disruption, which originated from AWS’s US-EAST-1 region, quickly spread beyond the United States, affecting users across Europe, Asia, and Africa. AWS confirmed that multiple services were experiencing “increased error rates and latencies,” adding that engineers were working urgently to mitigate the issue and determine the root cause.
The impact has been widespread, silencing Alexa devices and crippling apps and businesses that rely on AWS cloud servers, from Airtable and Perplexity AI to the McDonald’s app. Gaming platforms such as Fortnite, Roblox, and Rainbow Six Siege were also knocked offline, prompting a surge of complaints on social media. Downdetector recorded thousands of outage reports in the U.S. alone, while frustrated users on Reddit and X shared screenshots of failed logins and frozen dashboards. Perplexity CEO Aravind Srinivas confirmed the disruption, blaming the AWS outage for the platform’s downtime.
In a public update, Amazon said it would provide status reports every 45 minutes, but has not yet given a timeline for full restoration. The outage has also affected AWS’s own Support Center and Support API, leaving some organisations without access to troubleshooting services. The episode has renewed industry concerns over cloud dependency, as a large portion of the global internet relies on a small number of cloud providers—particularly AWS—to remain operational.
This latest outage adds to a growing history of disruptions tied to the US-EAST-1 region, including major incidents in December 2021, November 2020, and June 2023 that shut down platforms such as Netflix, Disney+, Zoom, Slack, and Twitch. As millions of users wait for services to return, the event has reignited debate over the risks of centralisation in global internet infrastructure and the vulnerability of essential services when a single cloud region fails.