Last updated: 21 October 2026, 2:40 AM PKT. Timeline and affected services verified against Google’s status page statements, UK parliamentary communications and multiple wire reports.
Quick answer: A power issue at a Google Cloud data centre near London knocked millions of customers offline on Tuesday, taking down banking apps including Lloyds, HSBC and Revolut, crashing the video platform England and Wales’ courts depend on, and prompting a Treasury Committee letter to the Bank of England and the FCA demanding answers. Google says the issue has been mitigated and fixed for most customers, with services restored by mid afternoon, and has promised an analysis of what went wrong.
Most people have never chosen Google Cloud, and on Tuesday they discovered it had chosen them. When a power failure inside one company’s London region can freeze your salary, your mortgage payment, and your day in court, the story stops being about technology and becomes about the invisible dependencies of modern life. This report explains what happened, who was hit, why one building can do this much damage, and the uncomfortable questions now being asked in Parliament.
The Timeline of a Morning Without Banks
Problems were first detected just after 7am on Tuesday, when monitoring picked up failures in Google Cloud’s europe-west2 region, the London based infrastructure that serves as the digital plumbing for a remarkable share of Britain’s financial and public services. Within minutes, the failures surfaced where consumers live. Revolut, the app based bank with millions of British customers, went dark, its users unable to log in, send money, or pay at tills where their cards were declined mid transaction. Lloyds and HSBC customers reported apps frozen and payments hanging. Halifax, Nationwide, NatWest and Bank of Scotland users joined the pile on within the hour.
Then the outage climbed the institutional ladder. HM Courts and Tribunals Service lost access to its video platform, knocking remote court hearings offline across large parts of England and Wales and drawing a furious note from the Criminal Bar Association about the justice backlog being hampered yet again. Royal Mail and courier DPD reported disruptions to tracking systems. By mid morning, the failure had its own segment on every news bulletin, and the Treasury Committee had written to the Governor of the Bank of England and the FCA chief demanding to know how the financial system’s resilience was left hostage to a single commercial provider’s power problem.
Google’s response followed the modern incident playbook, initial confirmation, a mitigation announcement before 1pm, restoration for most customers by 3pm, and a promise, repeated like a mantra, to publish an analysis of what went wrong. Revolut confirmed all customer money and data remained safe throughout, which is the sentence that keeps such events on the embarrassment page rather than the catastrophe page.
The outage in numbers
Start: just after 7am UK time, Tuesday 20 October
Epicentre: Google Cloud europe-west2 region, London
Restored for most: by 3pm UK time
Hit: Lloyds, HSBC, Revolut, Halifax, Nationwide, NatWest, Bank of Scotland, Barclays services, Royal Mail, DPD, HM Courts video platform
Response: Treasury Committee letter to Bank of England and FCA
Why One Building Can Freeze a Nation’s Money
The uncomfortable truth beneath Tuesday’s headlines is concentration. Banks did not hand their core banking to Google casually, they migrated workloads to cloud regions for flexibility and cost over fifteen years, and the big three providers, Google, AWS and Microsoft Azure, now sit under a startling share of the financial system’s digital operations. The regulators know this, the banks know this, and the resilience doctrine, redundant regions, failover capacity, multi cloud spread, exists precisely for days like Tuesday.
What failed, apparently, was not the doctrine but the willingness to pay for it fully. Redundancy across regions and providers is expensive, and every institution makes quiet calculations about how much resilience is enough. On Tuesday, the calculations of many institutions shared the same answer, and the same single point of failure. A power issue, in Google’s words, in one region, was enough to take payments, cards and courtrooms down together, which tells you the spare capacity was either absent, untested, or slower than the morning’s expectations.
The Committee’s letter will extract the honest answers in time, which banks ran which workloads where, why failover did not fire, and whether the regulators’ operational resilience rules, which have been tightening since 2023, were met in letter and failed in spirit.
The Human Cost of a Five Hour Outage
For the institutions, Tuesday was an engineering incident. For ordinary people, it was a day of small humiliations with real consequences. Parents could not load lunch cards for their children, as one furious customer put it. Workers’ card payments declined at petrol stations and grocers. Small businesses missed the morning’s takings. Court hearings, some involving remand decisions and family matters, simply did not happen, with defendants, victims and lawyers absorbing the delay. None of this registers on a status page, and all of it is why regulators treat operational resilience as a consumer protection issue, not an IT preference.
The psychological residue is subtler but real. Every outage of this kind teaches millions of people that the money in their app is a promise that depends on infrastructure they neither see nor chose, and the next time a provider stumbles, the queue at the branch and the cash machine will be longer. Trust in digital finance is built in decades and spent in mornings.
Google’s Bad Month, and the Cloud Concentration Question
The outage lands while Google is already defending its cloud reputation on two fronts. Earlier in the year, a separate cloud disruption had knocked large parts of the internet sideways, and Tuesday’s regional failure renewed the question of whether the industry’s pace of expansion is outstripping its operational maturity. Regulators in the UK want answers from Google about the outage, with the FCA and Bank of England expected to supervise the financial sector’s own follow up. Google’s promise to publish an analysis of what went wrong will be read in boardrooms far beyond Britain.
The structural question is bigger than one company. As AI drives unprecedented compute demand, the same hyperscalers are building faster than they can harden, and the economies that depend on them are consolidating dependencies faster than their regulators can map them. The UK’s experience this week, its courts and banks sharing a power strip, in effect, is a preview for every country that has migrated its institutions to three American companies’ basements. Germany and Japan have both studied the question; Pakistan’s own central bank and NADRA run hybrid architectures for exactly these reasons, a fact that looks wiser after every foreign headline like this one.
For readers who want to check service status themselves, Google publishes real time cloud health information on its official status dashboard.
What Happens Next
Three outcomes to watch. Google’s post incident analysis, promised publicly, the depth of which will signal whether this was a freak power failure or a design weakness. The FCA and Bank of England’s follow up to the Treasury Committee letter, and whether new operational resilience expectations name multi region requirements explicitly. And the quiet corporate aftermath, the boardroom reviews at every bank that went dark, each asking the same question, why did our failover not fire, and each privately pricing the answer.
The realistic prediction, based on how these cycles always run, is a season of resilience announcements, a few migrated workloads, some regulator letters, and a return to concentration by another name, because the economics that built the single points of failure remain the economics. Tuesday was a warning shot. The question is whether anyone was listening who can act on it.
We will update this page when Google publishes its analysis and when the parliamentary correspondence goes public.
A Field Guide: What To Do When the Banking Apps Die
Tuesday’s queues taught a practical lesson worth writing down, because there will be a next time. When payment apps fail, cash and cards on different rails behave differently, and the households that navigated the morning best were the ones with a small cash float, a card from a bank that runs on different infrastructure, and the habit of keeping one utility bill paid ahead rather than on the deadline. None of this requires paranoia, it requires the same redundancy thinking the banks skipped, priced for a family rather than a treasury.
For small businesses the checklist is sterner, because a dead card machine is dead revenue. Merchants who kept a manual receipt book and a float of change for exactly these mornings reported losing minutes instead of hours, and the petty cash discipline that feels old fashioned in normal weeks is the difference between trading and shutting the shutter. The five hour outage was, in effect, a free audit of every shop’s fallback plan, and the businesses that failed it now have the memo.
Households should also know their rights without needing them. Direct debits that fail on the provider’s side, payments stuck in limbo, and late fees triggered by an outage are recoverable, and the ombudsman framework exists precisely for the Tuesday morning when a bank’s app dies and a mortgage payment bounces. The banks’ goodwill gestures usually arrive first, but the entitlement exists whether or not the gesture does.
The Regulator’s Toolbox, Explained Simply
The letters to the Bank of England and the FCA invoke a framework most readers have never seen, the operational resilience rules that have required major financial firms to identify their important business services and prove they can keep running them through disruptions. The rules set impact tolerances, the maximum acceptable outage for payments, for account access, for trading, and Tuesday’s question, asked politely in parliamentary prose, is whether tolerances were breached and what that obliges anyone to do about it.
The toolbox available to regulators runs from information requests, through skilled person reviews, to fines and, in the extreme, restrictions on firms that cannot meet resilience expectations. None of that helps the parent whose lunch payment failed, which is why the more interesting lever is structural, the debate about naming cloud providers as critical third parties, a designation the United Kingdom has been circling since the last big outage, which would put the major cloud companies under direct regulatory supervision for the financial system’s sake rather than their own customers’ contracts.
That designation fight is the one to watch, because it decides whether the next outage letter goes to the banks or to the basement. The Treasury Committee’s correspondence will be published in time, and between now and then every institutional treasury in Britain is quietly re pricing its own concentration risk, the truest regulator of all.
Frequently Asked Questions
The Outage
What caused the Google Cloud outage in London?
Google attributed the outage to a power issue at its London based data centre serving the europe-west2 region, detected just after 7am UK time on 20 October, with services mitigated by 1pm and restored for most customers by 3pm. A full analysis has been promised.
Which services were affected?
Banking apps including Revolut, Lloyds, HSBC, Halifax, Nationwide, NatWest and Bank of Scotland, card payments, Royal Mail and DPD tracking systems, and the HM Courts and Tribunals video platform used for remote hearings across England and Wales.
Was customer money safe during the outage?
Revolut and the affected banks confirmed that customer money and data were safe, with the disruption affecting access and payments rather than balances or records.
The Aftermath
Why did one data centre affect so many banks?
Because many UK financial institutions run significant workloads in the same Google Cloud region, and their redundancy arrangements did not prevent simultaneous failures, a concentration risk regulators are now being pressed to examine.
What is the UK government doing about it?
The Treasury Committee chair has written to the Bank of England Governor and the FCA chief executive demanding answers about financial system resilience, and UK regulators have said they want details from Google about the failure.
Could this happen again?
Outages of this class recur across the cloud industry, which is why regulators treat concentration as a standing risk. Whether Tuesday’s specific failure repeats depends on Google’s fixes and on whether institutions pay for the full multi region resilience their regulators recommend.
