The AWS outage is back — and half the internet’s broken again

Amazon Web Services is facing one of its most disruptive outages in recent memory, causing widespread instability across major websites, apps, and online tools. What began as brief downtime overnight has turned into an ongoing issue that continues to affect users worldwide. From productivity platforms and games to streaming and payment systems, many services built on AWS remain unreliable. This live blog tracks the outage as it unfolds

Oct. 20 — 16:15 p.m. ET

According to the BBC, although the fault with the specific AWS location at fault is resolved (US-EAST-1 in northern Virginia), there are still issues affecting apps and websites that rely on AWS. There are particular problems with an AWS computing service called Lambda​​​​​​, which many services use to provide access to apps on smartphones, computers, smart TVs, and more.

Downdetector is still reporting widespread issues with numerous services, with Lyft and Hulu users particularly affected.

Oct. 20 — 12:35 p.m. ET

Amazon’s cloud services are flickering again. After a morning of instability that briefly seemed resolved, AWS is once more experiencing widespread outages. Many sites and apps that rely on Amazon’s infrastructure are loading slowly or not at all. We’re tracking the situation live as it develops.

Oct. 20 — 6:45 a.m. ET

As of 06:45 on October 20, it appears most of the AWS services are returning to normal operating status. Amazon says they’ve taken “additional steps to aid the recovery.”

Oct. 20 — 4:30 a.m. ET

It’s October 20, 2025, and you might be wondering why the internet suddenly feels broken. Some of your favorite apps aren’t working, websites refuse to load, and even major online services are timing out. For millions of people across the world, it’s not a local Wi-Fi issue — it’s a cloud one.

Popular platforms like Asana, Perplexity, Snapchat, and games such as Roblox, Fortnite, and Clash Royale have all gone offline or are struggling to connect. Streaming services, payment apps, and even chatbots are also affected. In other words, if it feels like half the internet just vanished, that’s not far from the truth. The full list of affected services is included below.

The source: a major AWS outage

The issue stems from the AWS (Amazon Web Service) outage. The AWS health report shows that there’s an outage in the US-East-1 region from “multiple servers.” This, in turn, seems to have caused a domino effect that took down the rest and led to this widespread outage. Amazon’s AWS is the backbone for many popular non-Amazon applications. So when AWS has a problem, companies like Netflix, Disney+, Venmo, and various other apps have historically experienced service disruption.

The outage is happening in the largest and most strategically important cloud region globally. This means that the outage is not contained to US East, but has a global ripple effect. The US-EAST-1 (N. Virginia) region is AWS’s oldest and largest data center hub. Crucially, the control planes for many global WS services, such as Amazon Identity and Access Management (IAM) and Amazon CloudFront, are housed here.

An issue in this region can therefore affect services worldwide, as applications need to authenticate or access the global control plane, even if their data resides elsewhere. So, as it turns out, if your app doesn’t use AWS save for 2FA authentication, then you can use it just fine if you’re still logged in. However, if you’re logged out, you will not be able to log back in during this outage, which is one of the many problems of 2FA and outsourcing it to other platforms.

What services are affected by the AWS outage?

A reminder of cloud fragility

Outages like this reveal how much of the modern internet rests on shared infrastructure. When one of these central pillars falters, the effects ripple globally and knock out everything from productivity tools to entertainment platforms.

For now, users can only wait for AWS engineers to restore service and for the web’s many dependent platforms to recover. It’s a rare but sobering reminder that even the cloud has single points of failure.