A power fault in Google Cloud's London region (europe-west2-b) knocked Revolut's card payments, ATM withdrawals and currency exchange offline for hours on 20 October 2026, with knock-on disruption reported at UK banks and ING Germany.
Revolut customers across Europe could not pay by card, withdraw cash or reach their accounts for much of Tuesday, after a power failure at a Google Cloud data centre in London knocked out infrastructure that a large part of the continent's digital economy depends on.
Google's own incident record is unusually specific. The disruption began at 23:08 Pacific time on Monday 19 October — 07:08 in London — and ended at 07:00 Pacific on Tuesday, a total of seven hours and 52 minutes. Google attributed the initial failure to a power event inside the europe-west2-b zone, the London region's second availability zone. Thirty-eight products were listed as affected, among them Compute Engine, Kubernetes Engine, Cloud SQL, Cloud Storage, BigQuery, Firestore and Cloud Load Balancing.
What went wrong, and when
The first public updates described only "elevated errors and intermittent network failures" with engineers still investigating. Google identified the power event in later bulletins, saying it had been "mitigated" and that engineers were prioritising the recovery of existing workloads. The company advised customers to migrate workloads to another zone or region where feasible — advice that is straightforward for a well-architected service and considerably harder for a bank's core systems.
Google said recovery of most workloads was visible by 05:10 Pacific and that the incident closed with only a small number of customers still affected. It has said it will publish a full analysis once its internal investigation is complete. That document, not Tuesday's status updates, is where the accountability will lie.
Revolut: money safe, services not
Revolut's problems began around 08:00 London time, roughly an hour after Google's incident opened. Users in the UK, Ireland, France, Spain, Romania and Germany reported failures. The company's status page showed card payments, top-ups by bank card, ATM withdrawals and currency exchange as unavailable or partly unavailable; its Spanish status page listed 31 functions as degraded, including transfers, stock trading and customer support.
Revolut said on X that "all our customers' money and data are safe." A person familiar with the matter told Reuters that the outage was tied to the Google Cloud disruption. The link is not incidental: Revolut runs its app and back-end systems on Google Cloud, and the two companies expanded their agreement in September 2025 into a multi-year deal that also covers the AI tools Revolut uses for fraud detection.
That is the uncomfortable detail. The same vendor relationship that gives Revolut its fraud defences also concentrates its operational risk in a single supplier and, on Tuesday, a single availability zone.
Wider fallout
Downdetector recorded spikes in user reports for HSBC, Barclays and Lloyds Bank, though it is not established that the Google outage caused them. ING Germany separately reported that some customers could not log in to online banking or its app because of a technical fault; Reuters did not link that incident to Google. Reports of problems with delivery and postal platforms also surfaced during the same window.
The regional pattern of the Revolut failures — UK, Ireland, France, Spain, Romania, Germany — is consistent with European customers routing through London infrastructure rather than with a global Revolut failure.
Why this matters
"The outage is a reminder of how deeply our economies depend on just a handful of cloud providers," Feng Li, associate dean for research and innovation at Bayes Business School, told CNN. "A problem in one region can interrupt banking, payments and other services well beyond the provider itself."
Two things stand out. First, the failure mode was mundane — a power event, not a cyberattack. Highly publicised attacks attract attention, but physical infrastructure faults at a single cloud zone cause broader and more immediate consumer harm. Second, the consolidation problem is structural and getting worse, not a one-off. Each multi-year cloud commitment signed by a digital bank deepens the dependency, and the mitigations available — multi-region failover, multi-vendor architecture — are expensive, slow to build and rarely prioritised until an incident forces the question.
Separately: Revolut's breach history
Tuesday's outage is not connected to Revolut's recent data security problems, and the two should not be conflated. In September, Revolut confirmed that it had disclosed sensitive customer information to an unauthorised party after receiving fraudulent data requests sent from a legitimate government email domain, apparently impersonating Italian authorities. Around 680 customers were affected, with exposed data reported to include identity documents and transaction records. Revolut subsequently said it would cover the cost of replacement identity documents for those customers, and the UK Information Commissioner's Office is investigating.
A group calling itself "iamnotavillain" publicly demanded roughly $3 million in Monero and threatened to sell stolen records; Revolut said it received no direct contact from the claimants. A second, unrelated incident affecting US trading customers through third-party broker DriveWealth was also reported in September, exposing names, emails and employment data.
The two stories converge on one point of governance: Revolut's exposure runs through third parties — a cloud provider for availability, brokers and email-based verification for security. Regulators and customers are entitled to ask which of those dependencies the company has actually stress-tested.
Sources
Google Cloud Service Health incident Xs2XTHzAVC9YnafnPyzj; Reuters via CNN; The Next Web; FTN News; Revolut system status; Financial Times reporting on the September breach; TechCrunch.