AWS outage takes down half the internet. Here’s how we fix it.

On October 20, 2025, a major AWS outage has taken down huge portions of the internet — including Amazon, Alexa, Snapchat, Fortnite, Roblox, Signal, Slack, Reddit, Canva, Airtable, ChatGPT, Perplexity, and Epic Games Store.

According to the AWS Health Dashboard(neues Fenster), multiple Amazon services are experiencing operational issues in the US-EAST-1 region (North Virginia), and the cause is still under investigation. It’s not the first time this happened, as similar AWS outages in the same region previously occurred in 2023, 2021, and 2020, each causing widespread service interruptions.

This is yet another reminder that the internet — once envisioned as a decentralized network where anyone could freely publish, connect, or build — has become dependent on a handful of corporations. When Amazon, Google, Microsoft, or Meta experience disruptions, the world goes offline.

Here’s why it matters and what we can do about it.

What is AWS?

Amazon Web Services (AWS) is Amazon’s cloud platform and the world’s largest provider of cloud infrastructure(neues Fenster) — the backbone that keeps much of the internet running. Rather than buy and maintain their own servers, companies rent computing power, storage, and databases from AWS.

For example, when you stream a movie on Netflix, send a message on Snapchat, or check an app for your government documents, the data is likely traveling through Amazon’s servers.

AWS’s scale and reliability made it the easy choice for millions of organizations, but that puts a lot of control in Amazon’s hands.

How the AWS outage exposes Big Tech’s risks

The issue with AWS — and Big Tech more broadly — is centralization. So much of the internet now depends on the infrastructure, rules, and resilience of just a few companies. And that dependence makes the system fragile and our privacy vulnerable. Here’s how:

Single points of failure: When a major platform goes down, it can impact millions of users across different services. For example, a Microsoft 365 and Azure outage in October 2025(neues Fenster) disrupted access to Teams, Outlook, and other business tools. Likewise, a Google Cloud outage in June, 2025(neues Fenster) took down services including Spotify and Discord.

Data concentration: When you use an app that runs on Big Tech’s cloud, your personal or business data likely lives on their infrastructure. The cloud provider and its partners may access your data for storage, processing, or analytics. The more data that flows through a few providers, the greater the risks of misuse, exposure, or loss of privacy.

Economic and political leverage: When so many companies and government agencies depend on a few providers, those providers gain enormous influence. They can raise prices, set terms, comply (or not) with government data requests, and dictate who gets to operate online — something we’ve seen repeatedly when Big Tech deplatforms apps or services overnight. This dependence also leaves Europe’s digital sovereignty at risk, as much of its infrastructure relies on companies based in foreign jurisdictions, with different laws and interests.

How to reduce the impact of an AWS outage

Here’s how individuals can stay connected and productive during future AWS disruptions:

Consider independent cloud storage: Privacy-focused providers like Proton Drive operate outside Big Tech infrastructure, offering end-to-end encryption and better resilience for sensitive files.

Keep offline backups: While AWS outages don’t cause data loss, they can temporarily block access to cloud-stored files. Keeping local copies of critical documents means you can still work during disruptions.

Use offline-capable tools: Choose apps that can function without an internet connection and sync automatically when connectivity returns. For example, Proton Drive’s desktop apps lets you keep editing files offline, then syncs your updates once the internet is restored.

If you’re a business, here are some practical ways to minimize downtime and protect operations during major outages:

Set up monitoring and alerts: Tools like AWS CloudWatch(neues Fenster) or third-party platforms such as Datadog(neues Fenster) can automatically detect performance issues and trigger backup systems early.

Use multi-region deployment: Businesses can host their applications or websites in multiple AWS regions (not just us-east-1, for example), so if one goes down, others stay online.

Implement failover and backup systems: Set up automated systems that redirect traffic or workloads to backup servers during outages. For instance, using content delivery networks (CDNs) like Cloudflare helps by caching content closer to users.

Develop and test a disaster recovery plan: Regularly rehearse how your team would restore critical operations if cloud services fail.

Building a more resilient internet

If the web is to remain a public good, we need to restore its diversity of infrastructure and ownership(neues Fenster). While we can’t rebuild the internet overnight, we can make conscious choices about the tools we use.

At Proton, we build privacy-first services that operate independently of Big Tech’s infrastructure. Our apps are open source and independently audited, which means anyone can verify how they work and that data stays private. Built in Switzerland and protected by strong privacy laws, Proton is designed for a digital world that’s private, secure, and free by default. We also support the growing Eurostack movement, which promotes European-built infrastructure and greater diversity in the cloud ecosystem.

When data is distributed — hosted by independent providers or controlled directly by users — no single company can control or disrupt it all. That’s how the internet was meant to work.