Google Cloud outage hits UK banks and courts after London data center power event

A power event struck Google Cloud’s data center in the europe-west2-b zone early on 20 October 2026, triggering outages that affected banking applications, payment processing and remote court services throughout the United Kingdom.

The incident began around 07:00 BST and lasted until most workloads recovered by mid-afternoon. Multiple Google Cloud products, including Compute Engine, Kubernetes Engine and Cloud Storage, recorded elevated error rates and intermittent connectivity failures.

Timeline and technical details of the power event

Google Cloud reported that the outage originated from a power event isolated to the europe-west2-b availability zone. Services experienced partial degradation starting at 23:08 PDT on 19 October, which translated to early morning hours in London. The company stated that mitigation steps restored the majority of affected workloads by 13:00–14:00 BST.

Engineers isolated the fault and rerouted traffic where possible. Residual issues lingered for a small number of customers into the afternoon, but the status dashboard showed broad recovery by late 20 October. Google Cloud committed to publishing a full post-incident report within days.

The outage coincided with normal business hours in Europe, amplifying visibility. Metrics published on the Google Cloud Service Health dashboard indicated spikes in error rates for Pub/Sub, Cloud Run and BigQuery during the peak period. No evidence of data loss or security compromise has been reported.

Impact on banking, fintech and public services

Revolut customers faced intermittent card payments, top-ups and ATM withdrawals, generating more than 1,500 reports on Downdetector. Lloyds Banking Group, HSBC and the Bank of Scotland experienced degraded online and mobile banking performance. Institutions confirmed that customer funds and account data remained secure throughout the disruption.

HM Courts and Tribunals Service recorded problems with its Cloud Video Platform, forcing some remote hearings in England and Wales to be delayed or adjourned. The Criminal Bar Association noted that proceedings in affected courtrooms could not proceed until connectivity stabilised.

Additional sectors felt secondary effects. Royal Mail and DPD reported tracking and booking delays, while Rightmove and budgeting app Emma saw service interruptions. A limited number of NHS app users encountered login issues. Most reported problems remained confined to the UK market.

Recovery status and operational implications

By late afternoon on 20 October, Google Cloud confirmed that the power event had been contained and that service levels had returned to normal for the overwhelming majority of customers. Remaining affected workloads continued to be addressed on a case-by-case basis.

The episode underscores the concentration risk that arises when multiple critical services rely on a single cloud availability zone. Financial institutions and government bodies that use Google Cloud for production workloads may review their multi-region architectures and failover procedures in the coming weeks.

Enterprise teams should examine incident reports once released and test documented runbooks for similar power or network events. Regular validation of backup connectivity and alternative providers remains a standard practice for regulated industries that depend on cloud infrastructure.