The Hyperscaler Trap: How a Single Google Datacentre Glitch Ground the UK to a Halt
A localized power outage inside Google’s europe-west2-b zone was all it took to trigger a cascade of systemic disruptions across the United Kingdom. On a single morning, major banking applications stalled, remote court hearings across England and Wales were suspended, and patients encountered sudden glitches on the NHS app. While Google maintains more than a dozen distinct zones across Europe to service its clients, the failure of this single London facility instantly paralyzed core services, exposing how profoundly modern operations rely on a narrow set of foreign corporate platforms.
When Essential Services Hit a Wall
According to incident tracking on Downdetector, the disruptions surged to a peak around 9:00 AM before achieving partial stability by mid-afternoon. The damage was concentrated across vital commercial and public nodes. Financial institutions bore immediate friction: Revolut confirmed widespread service disruption while moving to assure clients their capital and data were safe, while Lloyds acknowledged operational snags across its digital channels alongside outages impacting HSBC and Royal Mail.
The operational breakdown went well beyond private commerce. The HM Courts and Tribunal Service, which coordinates remote proceedings, saw its Cloud Video Platform disabled. Because Google Cloud stores data and executes fundamental workloads for numerous third-party organizations, an isolated electrical failure instantly compromised judicial hearings and state-administered services.
A Pattern of Fragile Consolidation
This disruption is not an isolated event, but part of a persistent structural pattern. A year prior, an outage at Amazon Web Services’ us-east-1 node in Virginia—the highest-density cluster of datacentres on Earth—triggered worldwide operational paralysis. That incident lasted nearly half a day, disrupting Signal, disconnecting Roblox, and overheating smart mattresses globally.
As internet infrastructure expert Michał “rysiek” Woźniak pointed out, hyperscale cloud operators have reached a point where they are effectively "too big to not fail." More critically, they have embedded themselves so deeply into foundational institutional workflows that departing from them is treated as an organizational impossibility. What is sold as limitless resilience is regularly betrayed by the physical, fragile reality of regional power grids and server racks.
The Sovereign Cost of Cloud Monopoly
The broader crisis is one of concentrated dependency. Up to 60% of the UK government’s digital infrastructure is currently hosted on cloud services dominated by three firms: Amazon, Microsoft, and Google. This public sector reliance persists despite acute awareness of the liabilities involved. British police, for instance, retain critical records on Microsoft’s Azure cloud despite explicit recognition that this setup creates systemic exposure to foreign actors, foreign intelligence access, and recurring cloud outages.
When state agencies and critical consumer institutions outsource their backbone to an oligopoly of hyperscalers, single points of failure become national choke points. As routine datacentre disruptions cascade through justice systems and healthcare portals, the convenience of centralized hosting increasingly trades off against basic systemic security.