#itleadership — Public Fediverse posts

Thursday's Microsoft 365 outage took down more than email; it also affected the AI tools plugged into it, and the outage pushed back one public company's earnings date.

Outlook, Teams, and other Microsoft 365 apps were disrupted for more than five hours. Reuters reports that Outlook problem reports peaked at 138,058 on Downdetector.

Microsoft said several servers were unexpectedly unavailable and that it wasn't a traffic spike. The outage also hit the Microsoft connectors used by OpenAI's and Anthropic's AI services, and the Workhorse Group moved its third-quarter results from November 12th to November 16th because of the disruption. 😬

Microsoft, a core internet cloud provider, had a bad afternoon, and it traveled through email, chat, AI tools, and an earnings calendar. And in my experience, most continuity plans can't cover an outage like this. Since these #SaaS providers are essentially utilities, it's like having the power go out in your home.

If email and Teams go down together, how do you reach your people? Which of your AI tools and automations stop when Microsoft 365 does? And who owns that list? What did Thursday break for you that you didn't expect?

https://www.reuters.com/technology/microsoft-365-outlook-down-thousands-us-users-downdetector-shows-2026-11-12/

My status board monitors many SaaS platforms, and on Thursday evening, 13 of them were reporting trouble. HubSpot, Zoom, DocuSign, Monday.com, Qualtrics, KnowBe4, OpenAI, and Anthropic all named Microsoft in their own notices. 😬 You can check it anytime:
https://briangreenberg.net/service-status

#Microsoft365 #BusinessContinuity #CloudOutage #ITLeadership #CIO #security #privacy #cloud #infosec #Microsoft365

It's #AI, and the question we need to be asking is... "Who's watching the agents?"

That was the question on the table Tuesday evening at ELEVATE - IT Leaders & Executives' roundtable event at Zoro.com's Chicago headquarters. And in my experience, most companies can't answer it yet.

Dimple Thakkar led the discussion on protecting enterprise data now that AI agents are doing real work. Brian Ivanov and the Zoro team hosted, and Tom Parker keeps putting together good rooms. Thank you, all three.

My own answer hasn't changed much. An AI agent is a new kind of employee with a browser, API keys, and a service account. In plenty of shops, the main thing holding it back is just a line in a prompt. 😬 You wouldn't hire a person that way.

So, ask the boring questions... What can each agent touch with the credentials it holds today? Who cut the access it doesn't need? Where does its activity get logged, and can it edit that log? Has anyone tried to talk it past its limits before an attacker does? And who owns those answers? If you can't name that person, nobody's watching.

How is your organization handling it? Does agent oversight sit with security, IT, or the business?

https://www.alinedconsulting.com/elevate

#AIGovernance #AgenticAI #cybersecurity #CIO #ITLeadership #security #PrivacyMatters #cloudcomputing #infosec #AlinedConsultingGroup