A critical Microsoft outage lasting more than 22 hours has raised questions about how heavily vital services rely on a handful of major tech firms.
The disruption, which affected Microsoft 365 services such as Outlook, was caused by routine automated checks that took some servers offline, overloading the remaining ones with traffic.
The recovery took nearly a full day while engineers gradually restored services, highlighting how embedded the tech giant's services are in modern society.
Flora Hamilton, executive director of the Small Business Charter, said the outage was a "stark reminder" of how much firms "rely on digital systems" - like Microsoft Outlook - "to keep their operations running".
"For many small and medium-sized enterprises (SMEs), access to email is fundamental to everything from communicating with customers and suppliers to generating income," she said.
The disruption was not just confined to businesses, but also affected organisations and public institutions around the world.
Around 430 million people globally use Microsoft 365, and issues were reported across more than 70 countries.
The fallout from the outage had a wide-ranging effect , knocking out email systems for the US House of Representatives and Australian intelligence agencies, as well as disrupting court systems in Singapore and airline support desks in India.
Microsoft has since apologised for the issue, which it said was due to a technical failure rather than a cyber-attack.
Walter Pasquarelli, an AI strategist, told the BBC that for governments, businesses and critical infrastructure, the outage had "put their vulnerabilities and dependence in plain sight".
"Looking ahead, as AI assistants are plugged into our email, calendars and files, a failure like this may spread even further," he said.
Gergely Orosz, a former leading software engineer at several major tech companies, said while these kinds of outages weren't unheard of, what was unusual was the time it took for Microsoft to recover from it.
A spokesperson for the US tech company said "given the scale of demand" it restored services gradually.
"While this took longer, it helped ensure a stable restoration," a spokesperson said.
Orosz said the resulting upset showed "our reliance on a few large US corporations".
He said a solution for the future could be to split services per region to reduce the impact, something which already happens with Amazon's cloud computing service, AWS.
For now, Microsoft said it was reviewing what went wrong, with the findings expected to inform steps to prevent a similar disruption from happening again.
