One Power Event in London Knocked Revolut, Lloyds and HSBC Offline as Google Cloud Restores Its Biggest UK Region

A single power failure inside one availability zone of Google Cloud's London region briefly broke a surprising amount of British digital life on Tuesday. Google reported elevated errors and intermittent network failures in europe-west2-b, part of its europe-west2 region, starting around 07:08 London time, citing what it called a power event at a third-party facility.

The blast radius was wide. Affected Google Cloud services included Google Kubernetes Engine, Cloud Run, App Engine, Cloud Filestore, BigQuery Data Transfer Service, NetApp Volumes and Managed Service for Apache Kafka. Downstream, digital bank Revolut saw card payments, transfers, ATM withdrawals and trading disrupted; Lloyds Banking Group reported problems across Lloyds, Halifax and Bank of Scotland online services; HSBC acknowledged access issues. Royal Mail, courier DPD, property portal Rightmove, the NHS app and the courts' remote-hearing video platform also logged trouble.

Revolut, which runs as a cloud-native bank with no legacy fallback stack, was among the hardest hit, with more than 1,500 user reports on outage trackers shortly after 08:00. The company said customer money and data remained safe throughout and apologized for the disruption. UK regulators took note: the Financial Conduct Authority said it was engaging with Google and affected firms to assess the impact, with no enforcement action announced.

Google said the power event was mitigated within hours and prioritized recovery of existing workloads, initially targeting restoration around 13:00 UTC. By Tuesday evening a spokesperson confirmed services had been restored, and the company promised to publish a post-incident analysis on its service health dashboard once an internal investigation completes.

The episode is a case study in concentration risk. High availability across zones is not a default; it is something customers pay to engineer, and many institutions keep critical services, identity checks, SMS gateways and risk systems chained to the same region. The same region suffered a cooling-system failure in July 2022 that interrupted service for more than 42 hours. And the incident landed almost exactly one year after an outage at AWS's largest cloud site knocked more than 1,000 sites offline, including UK banks.

For an AI industry pouring hundreds of billions into cloud capacity, the message is uncomfortable: the infrastructure layer is now critical national infrastructure, and a single-zone electrical fault can freeze payments for millions of people. Expect UK regulators, who have already designated major cloud providers as critical third parties for finance, to press harder on multi-region resilience requirements in the post-incident review.