Microsoft's Azure cloud platform suffered a major outage in its California region yesterday, lasting almost five hours after a maintenance error severed fibre connections. The incident, which the company attributed to a mistake during routine maintenance, took 27 services offline and left customers unable to access critical cloud resources.
The disruption began early on 23 July 2026 and affected a wide range of Azure services, including virtual machines, storage, and networking capabilities. Microsoft confirmed the root cause was a fibre cut during planned maintenance work, though it has not yet disclosed full details of the error or which specific services were hardest hit. The company said it had implemented a fix and services were restored by the afternoon.
For UK businesses, the outage serves as a stark reminder of the risks tied to geographic concentration of cloud infrastructure. Many British firms rely on Azure for data storage, application hosting, and disaster recovery, often with little visibility into the physical resilience of the data centres they depend on. Dr. Eleanor Marsh, a cloud computing researcher at the University of Cambridge, said: 'Single-region failures like this demonstrate why redundancy across multiple regions is not just a technical recommendation but a business continuity necessity. UK organisations should audit their cloud dependencies and ensure they are not overly reliant on a single provider or location.'
The outage also raises questions about regulatory oversight. The UK's Information Commissioner's Office (ICO) requires organisations to ensure the security and availability of personal data under the UK GDPR. A prolonged cloud failure could lead to data loss or breach of contractual obligations, potentially triggering enforcement action. Meanwhile, the EU's AI Act, which applies to UK companies operating in Europe, also mandates robust infrastructure for high-risk AI systems—many of which are hosted on platforms like Azure.
For the UK economy, the incident underscores a broader vulnerability: as businesses increasingly migrate to the cloud for everything from payroll to AI workloads, a single provider's error can cascade across sectors. The cost of downtime for a mid-sized enterprise can run into tens of thousands of pounds per hour, not counting reputational damage. The outage also reignites debate about the dominance of a handful of hyperscale cloud providers—Microsoft, Amazon Web Services, and Google Cloud—and whether UK regulators should push for greater interoperability and failover options.
Microsoft has not yet indicated whether it will offer service credits or compensation to affected customers. The company said it would conduct a full post-incident review and share lessons learned. For now, UK IT leaders are being urged to review their cloud architecture and ensure they have robust backup plans in place.