Google's cloud had a very bad morning in Britain. A "power event" in its europe-west2-b zone — one of three zones in its London cloud region — began at 07:08 BST and spiraled into elevated errors and intermittent network failures across 38 cloud products in the wider region, according to Reuters. By late afternoon Google said full service was restored. In between, much of Britain's digital plumbing broke at once.

Revolut was the worst of it. The digital bank's customers reported cards declined at tills and error messages at login, with card payments and ATM withdrawals hit — a rough morning for its 13 million UK customers, as Beyond the Basket reported. Revolut confirmed the disruption and said customers' money and data are safe, but DownDetector was soon a wall of spikes: Lloyds, HSBC, Barclays, Halifax, NatWest, Nationwide and Bank of Scotland all reported problems, alongside Royal Mail, DPD and Rightmove.

The blast radius went further than banking. DownDetector climbed for NHS England, HM Courts and Tribunals Service and GOV.UK — and the Criminal Bar Association said the Cloud Video Platform, used for remote court hearings, was down across large parts of England and Wales from 10am. Courtrooms went dark because a power cable did.

The timing is almost too neat: this lands exactly a year after Amazon Web Services went down on 20 October 2025, taking Lloyds, Snapchat, Ring and the government's Gateway login offline with it. Two Octobers, two clouds, nearly identical damage. That pattern is why the outage matters more as policy than as plumbing.

In July, the UK brought Google Cloud, Microsoft, Amazon Web Services and Oracle under direct regulatory oversight as critical third-party suppliers to the financial sector — precisely because an outage at one provider can now disable multiple firms at once. The EU has designated Google Cloud among its critical providers too. The Financial Conduct Authority said it was "aware of an incident affecting services provided by Google and [was] engaging with them and affected firms to understand the impact." Translation: the stress test the regulators designed for is now running live.

Google's own response undercuts the redundancy story clouds like to sell. For the first two and a half hours there was no workaround; the eventual guidance was to move affected workloads to another zone or region "if feasible." Multi-zone failover is a design choice, not an accident — and it costs money, so plenty of customers hadn't made it. When the failover works, the cloud is resilient. When it doesn't, one zone's power event becomes a national outage, and the customer is the one doing the migrating.

The takeaway isn't that clouds fail — they do, regularly. It's that the UK's financial system now routes its cards, its deliveries and its court hearings through a handful of providers the state has formally declared too important to fail, and failover remains an optional extra. Until that math changes, expect the third Thursday in October to keep feeling like a tradition.

— via Reuters