Google Cloud Outage Knocks UK Banks Offline — Exposes Systemic Risk

A power failure at a Google Cloud datacentre in London on Tuesday caused a significant outage, knocking critical services offline for UK banks including Revolut, Lloyds, and HSBC. The incident left millions of customers temporarily unable to access their money and stalled judicial hearings, exposing the fragility of national infrastructure built atop a small number of technology providers.

Google apologized for the disruption, which it attributed to a “power event” within its “europe-west2-b zone,” according to The Guardian. The company stated that services had been restored for “most” users. However, the impact was immediate and widespread. Revolut was among the hardest hit, the BBC reported, with its app and card services going down for many customers. Patrons of Lloyds and HSBC also reported being unable to make payments or transfer funds.

Cascading Failure Hits Banks and Courts

The disruption was not limited to the financial sector. The Guardian also noted that the outage stalled court video cameras, putting some legal proceedings on hold. Patients even reported temporary glitches with the NHS app. This cascade of failures from a single physical event demonstrates how deeply integrated cloud services have become in daily public and private life.

The pattern is a familiar one. A localized hardware problem—in this case, a power issue at a single datacentre—creates a single point of failure that ripples outwards, affecting dozens of seemingly unrelated services that all share the same underlying infrastructure. While cloud providers build in redundancy, this event shows that a zone-level failure can still have significant consequences.

'Too Big to Not Fail'

This outage is less a story about a technical glitch and more about market concentration. As one expert told The Guardian, the reliance on a handful of key tech companies leaves the UK vulnerable. The phrase “too big to fail” is common in banking, but for cloud infrastructure, the reality may be that these providers are “too big to not fail.” When a single company’s infrastructure supports a significant portion of a country's banking, government, and healthcare services, its failures become national-level events.

The consensus from multiple reports is that while Google has resolved the immediate issue, the underlying problem remains. The convenience and power of centralized cloud platforms come with the inherent risk of centralized failure points. This incident serves as a mandatory, and for many, costly, reminder that “the cloud” is just a set of physical buildings, and sometimes the power goes out.