A Microsoft outage lasting nearly 17 hours across Britain and Europe has paralysed the UK government and the courts system, taking down Outlook, Teams and Copilot and forcing officials to activate contingency plans. The disruption has hit HMRC, the Department for Digital, Culture, Media & Sport (DCMS), the Courts and Tribunals Judiciary, four police forces and NHS online services, while Microsoft has ruled out a cyber attack as the cause.

The immediate institutional cost falls on the courts. Civil court judgments are normally handed down remotely by email before publication on the Bailii database and the Judiciary website, but most government departments, including the Courts and Tribunals Judiciary, run their email on Microsoft's Outlook platform. City AM understands officials are assembling a plan so that journalists and parties can access legal decisions, but the outage could push some judgments back until Monday.

What caused the failure?

Microsoft attributed the disruption to servers becoming unexpectedly unavailable, which placed additional strain on its remaining infrastructure. The company said its repair attempts produced only partial results and warned that services could become unresponsive again even after restoration.

"We're continuing to work on recovering our frontend servers while implementing further traffic optimisation actions to improve service stability and increase capacity," Microsoft said in a statement. "We're aggressively pursuing all pathways to reduce impact."

The company has yet to provide a timetable for full recovery.

Which parts of the state went dark?

The outage exposed how deeply central government depends on a single US vendor:

• HMRC confirmed its emails were down, though a spokesperson said "phone lines are working as expected" and encouraged people to use online services if they need support. "We will continue to assess any impacts on our customers," they added.
• DCMS staff are understood to be blocked from their emails.
• The Metropolitan Police, Dorset Police, South Wales Police and Thames Valley Police all reported disruption. The Met warned that online crime reports may not reach officers and advised the public to call 101 for urgent non-emergency matters or 999 in an emergency.
• NHS services reported problems accessing Rapid Health, an online consultation platform used by GP practices.
• The Cabinet Office escaped the outage because it uses Google rather than Microsoft.

Rachel Harrison, national secretary at the GMB union, told Sky News the outage was having a "serious impact" across ambulance and NHS trusts. "Those who work in 111 services are unable to refer patient details, which is creating a massive backlog and has the potential for chaos," she said.

Defence minister Luke Pollard confirmed that government engineers were working with Microsoft to assess the disruption. "We've already been in touch with Microsoft and the dialogue between our cyber resilience teams across government and Microsoft continues," he told Times Radio. "I want to see this fixed; it is on Microsoft to do so."

A government spokesperson said "robust contingency plans" were in place to minimise disruption.

How far did the damage spread?

The failure reached beyond the British state. Swiss luxury goods group Richemont reportedly delayed the publication of its half-year results, which had been due on Friday morning. AI companies OpenAI and Anthropic both reported problems with services connected to Microsoft's infrastructure.

The timing sharpens the political edge of the incident. It comes just days after the UK's National Cyber Security Centre (NCSC), the government's cyber defence agency, urged businesses to strengthen their resilience against technology failures and supply chain disruption.

Academics say the episode demonstrates the concentration risk embedded in global corporate IT. Toby Murray, a computing professor at the University of Melbourne, warned of the "inherent dangers" of businesses relying on a handful of technology giants for critical infrastructure. An outage at a single provider, he said, could "dramatically reduce business productivity globally".

What happens next?

Microsoft said it was continuing to work on restoring services. Its warning that recovered platforms could degrade again means departments cannot yet treat the incident as closed.

For the courts, the practical deadline is Monday, by which time delayed judgments would need to reach parties and media if Friday's remote hand-down proved impossible. For ministers, the question posed by a 17-hour, multi-agency failure of one vendor's software will outlast the outage itself.