BTC $68,479 
EQQQQ $655.75 Last updated 8 Nov 2026, 12:03 UTC
ETH $1,926 
MGLOEURO $111,813 
USDT $0.9993 
BNB $630.36 
USDC $0.9995 
XRP $1.01 
SOL $79.51 
TRX $0.3130 Technology · Explained
Microsoft says Outlook outage is not a cyber attack as recovery continues
Microsoft blamed overwhelmed connection-handling software for the multi-day disruption, which also cut ChatGPT and Claude off from users' Microsoft 365 data.
Key takeaways
- Cause. Microsoft pointed to a core process on affected servers becoming overwhelmed, causing capacity loss, and said it found no sign of malicious activity.
- Knock-on effects. ChatGPT and Claude lost access to Outlook mail, calendars, Teams chats and SharePoint files because their connectors rely on Microsoft's servers.
- Recovery. Microsoft reported broader improvements in North America, sustained gains in much of Europe and encouraging signals in Asia Pacific, while warning of intermittent issues.
What happened
Microsoft's Outlook, Teams and Copilot services remained disrupted on Friday after an outage that began on Thursday, with the company saying it had found no indication of malicious activity behind the failure. To ease the backlog, Microsoft extended an email timeout by 48 hours so that messages held up during the incident could be delivered as systems came back online.
In updates cited by Protos, the company said a core process on affected servers had become overwhelmed, producing a loss of capacity, and that the measures taken so far had not brought the relief it anticipated. Microsoft described the trigger as elevated load within its infrastructure.
The company said it was seeing wider improvements in North America, continued progress across much of Europe and encouraging recovery signals in Asia Pacific. It cautioned that some users could still run into intermittent problems while it prioritized clearing email queues and reducing the backlog.
The effects reached beyond Microsoft's own applications. OpenAI's status page showed elevated errors for its Microsoft 365 connector, warning that new connections might fail and tools might not work properly, while Anthropic reported similar trouble for Claude, including its Claude for Outlook add-in, and linked the issues to Microsoft's incident. The connectors are what let the assistants read Outlook email, calendars, Teams chats and SharePoint files, so the integrations faltered once Microsoft's servers did.
Why it matters
Microsoft 365 sits underneath a large share of everyday workplace communication, and Protos reported that the affected services serve 430 million Microsoft users. When a single provider's authentication and connection systems stumble, the consequences spread to third-party products built on top of them, which is what happened to the AI assistants. The incident also illustrates how quickly speculation about cyber attacks arises when major platforms fail, even where the vendor points to internal capacity problems rather than an intrusion.
What the data shows
Among the figures reported around the incident: 430 million Microsoft users affected by the service disruption, a 48-hour extension of the email timeout intended to let delayed messages catch up, a 15% decline in Nvidia shares over five trading days, and a bitcoin price of $68,000 cited by Protos. Microsoft's public statements did not include a timetable for full restoration.
Background
The disruption capped a turbulent week for the technology sector. Protos reported that Nvidia's former CEO Jensen Huang resigned following a series of exposed extramarital affairs, that Nvidia's stock had fallen 15% over five days, and that the slide appeared to weigh on wider financial markets, including bitcoin, which fell to $68,000. Separately, outages had hit Japanese credit unions and banks in the UK and Australia in recent weeks; those were likewise not attributed to malicious activity, unlike the cyberattacks that targeted European supermarket fridges.
What is still unclear
Several details remain unresolved. Microsoft has not said when capacity will be fully restored or how large the residual email backlog is, and its descriptions of the root cause have stayed at the level of an overwhelmed core process and elevated load. Anthropic listed claude.ai as showing degraded performance without saying whether that was connected to the Microsoft incident, even as it described its API and Claude Code as operating normally. It is also not yet clear how quickly the affected AI connectors will return to full service once Microsoft's systems stabilize.
What to watch
The practical markers to follow are the status pages maintained by Microsoft, OpenAI and Anthropic, and whether Microsoft publishes a fuller post-incident accounting of the failure. Also worth watching is how fast the queued email backlog clears in the regions Microsoft flagged as still recovering, and whether other services that depend on Microsoft 365 connections report fresh problems as traffic returns to normal.
Questions readers ask
Was the Microsoft outage caused by a cyber attack?
Microsoft attributed the disruption to elevated load and an overwhelmed core process on affected servers, and said it had found no evidence of malicious activity.
Why did ChatGPT and Claude lose access to email and calendars?
Both assistants rely on Microsoft 365 connectors to read Outlook email, calendars, Teams chats and SharePoint files. Those connectors stopped working when Microsoft's servers failed, and OpenAI and Anthropic each reported errors tied to the Microsoft incident.
Is the outage over?
Microsoft reported continued recovery across affected regions and said some users might still see intermittent issues while it works through email queues and backlogs. The company has not given a full restoration timeline.
