When AWS suffered a major outage, many were surprised to learn that Signal relies on Amazon’s cloud. Signal’s president Meredith Whittaker explains why that surprise reveals a deeper misunderstanding of how today’s internet actually works.
A debate that shows how little we understand internet infrastructure
When AWS suffered widespread outages last week, some Signal users were surprised to learn that the privacy-focused messaging app partially relies on Amazon’s cloud infrastructure. Meredith Whittaker, president of Signal, took to social media to offer clarification and provide broader context to the ongoing debate.
In a thread, Whittaker explained that despite the use of AWS, user privacy is still preserved. “We use encryption to make sure no one but you — not AWS, not Signal, not anyone — can access your communications”, she wrote.
But for Whittaker, the surprise itself was troubling: it revealed just how little awareness there is about the concentration of power among the world’s major cloud providers.
“It’s concerning,” she noted, “because it indicates that the extent of the concentration of power in the hands of a few hyperscalers is way less widely understood than I’d assumed.”
It’s not about Signal’s choice — It’s about the lack of alternatives
Whittaker urged critics to look beyond Signal’s specific setup and instead question why nearly every global, real-time communications platform — from Signal to X, from Palantir to Mastodon — relies in some capacity on AWS, Microsoft Azure, or Google Cloud.
Running a platform like Signal, she explained, isn’t as simple as “renting a server.” It demands a “pre-built, planet-spanning network of compute, storage, and edge presence” that ensures low latency and high availability for millions of concurrent users — whether they’re in Cairo or Cape Town, Bangkok or Berlin. This kind of infrastructure requires billions of dollars in investment and continuous maintenance.
Even if Signal somehow had those billions, it would still face another barrier: expertise. The operational knowledge, tooling, and reliability practices that underpin modern internet infrastructure were developed within the hyperscalers themselves. “The very language of modern SRE [Site Reliability Engineering] came out of these companies,” Whittaker wrote, “and is now synonymous with ‘the cloud.’”
Dependency is not just in the cloud
Signal’s reliance on AWS, Whittaker pointed out, is only one layer of a broader dependency stack that touches nearly every device and operating system. The app runs on iOS (Apple) and Android (Google) for mobile users, and on Windows (Microsoft) for desktops. Each of these ecosystems represents another form of concentrated power.
“The problem is the concentration of power in the infrastructure space that means there isn’t really another choice: the entire stack, practically speaking, is owned by 3–4 players.”
Learning moment for everyone
Despite those constraints, Whittaker emphasised that Signal’s architecture still preserves its integrity and privacy guarantees. End-to-end encryption ensures that, even when data moves across AWS infrastructure, no one but the sender and recipient can read it.
Her closing message: that the AWS outage and following discussion could be a “learning moment” — one that highlights the fragility of our digital ecosystem and the dangers of centralising the world’s “nervous system” in a handful of corporate hands.
📣THREAD: It’s surprising to me that so many people were surprised to learn that Signal runs partly on AWS (something we can do because we use encryption to make sure no one but you–not AWS, not Signal, not anyone–can access your comms). It’s also concerning. 1/
— Meredith Whittaker (@meredithmeredith.bsky.social) 27 oktober 2025 om 11:38
[image or embed]
