The Microsoft Outage Also Hit ChatGPT and Claude. But There’s an Easy Workaround

A massive network failure at Microsoft disrupted major enterprise services, Microsoft 365 applications, and third-party AI tools including OpenAI's ChatGPT and Anthropic's Claude.

Reported by outlets including Inc., CNN Business, and The Next Web, the disruption began on Thursday, November 12, 2026, dragged into Friday, November 13, and resolved around Saturday, November 14, 2026. The widespread technical failure impacted 74 countries and territories in barely an hour, affecting services relied upon by more than 430 million Microsoft 365 users worldwide.

What caused the Microsoft outage?

Microsoft stated that the disruption stemmed from an issue during an earlier network capacity upgrade that surfaced following a recent network change. According to Open Magazine, software establishing secure connections on affected servers became overwhelmed as lost capacity placed intense strain on remaining machines.

While Microsoft maintained the issue was internal, differing descriptions emerged from public officials. Australia’s home affairs minister Tony Burke described the event to reporters in Sydney as an update that had gone wrong, likening it to the July 2024 CrowdStrike incident. However, Microsoft officially clarified that while the 2024 CrowdStrike failure occurred on client PCs, the November 2026 failure occurred directly on Microsoft's servers. Microsoft CEO Satya Nadella's team and a Microsoft spokesperson confirmed to The Guardian that there was no indication of a cyberattack.

Which services and regions were affected?

Outlook problem reports peaked at roughly 138,058 at the US peak on Thursday, November 12, around 1:00 p.m. US Eastern time, before falling to about 34,490 by the US evening. OpenAI posted its first update regarding downstream AI impacts on Thursday at 1:29 p.m., while Anthropic's incident began at 17:52 UTC on Thursday.

The global footprint of the failure touched numerous sectors and locations, including:

• North America: Colorado (Arapahoe County Sheriff’s Office), Jacksonville, Austin, Dallas, Miami Beach, and Greensboro.
• Asia-Pacific & Europe: India, Australia (including Service NSW, federal police, and a national spy agency), Hong Kong (Morgan Stanley, Goldman Sachs, Standard Chartered), Japan, the Philippines Supreme Court, and the UK.
• Corporate Entities: Workhorse, Morgan Stanley, Goldman Sachs, Standard Chartered, ING Australia, IndiGo, and Akasa.

What are the financial stakes?

An initial estimate by the Anderson Economic Group placed the cost of lost productivity at least at $7 billion. Patrick Anderson, president of the Anderson Economic Group, estimated that roughly 26 million users on Thursday and 51 million users on Friday experienced operations disrupted in a way that caused an economic loss. Furthermore, Anderson estimated that Microsoft could conceivably need to refund customers about $800 million a day in subscription fees.

Microsoft issued a public apology, with a statement on X reading: “We are deeply sorry for the disruption this incident has caused our customers and partners. We know the trust they place in our services and take that responsibility extremely seriously.” Charles Lamanna, Microsoft’s executive vice president for Copilot, Agents and Platform, also posted an apology on X confirming no malicious activity.

What happens next for users and administrators?

Microsoft stated it will release a full post-incident report after completing its internal review. Microsoft 365 administrators can track the wider incident under identifier MO1550821 and the mailbox incident under EX1550543.

In the wake of the technical failure, authorities and financial institutions—including Standard Chartered and ING Australia—issued warnings regarding secondary threats, specifically scammers impersonating technical support staff to exploit the confusion.

This is educational information, not legal, financial, tax, or investment advice.