Who Else Has the Keys to Your Microsoft 365 Environment? 

Blog Andrew Hickman todayNovember 19, 2026

Background
share close

The ConnectWise Cloud Backup incident is a reminder that the applications we authorize can become some of the most powerful identities in our businesses. 

Organizations have spent years strengthening employee sign-ins with multifactor authentication (MFA), suspicious-login alerts, and password training. That’s important work. But here’s a question I rarely hear from business leaders: 

Who can access your company’s information without an employee signing in at all? 

Think about your backup provider, document management platform, and security software. They may connect to Microsoft 365 around the clock, with permission to read or change enormous amounts of information. 

The October 2026 ConnectWise Cloud Backup incident showed what happens when that trust is abused. Attackers used compromised vendor application credentials to access customer Microsoft 365 environments. Investigators observed attempts to elevate privileges and add credentials to other applications. Microsoft attributed the campaign to a China-aligned group it tracks as Storm-0940. Successful authentication alone doesn’t establish that customer data was stolen. 

I covered the technical details in this week’s episode of The SOC Brief. The broader question for businesses is worth its own discussion: Do you know how much access you’ve given outside applications, and would anyone notice if that access were misused? 

The Other Identities in Your Business 

When most people hear “identity security,” they think about employee accounts. But modern cloud environments contain identities for software, too. 

Microsoft calls one common type a service principal. In plain English, it’s the identity an application uses to access resources you’ve authorized it to use. A backup tool, for example, may need to read mailboxes and files every night without asking a person to approve each operation. 

That’s how many legitimate services work. 

The risk appears when an application has powerful permissions and somebody else gains control of the credentials that represent it. The system may recognize the application as trusted because an administrator approved it months or years ago. Employee MFA protections don’t necessarily apply to that application-only sign-in; no employee is being asked to log in. 

So strong employee MFA doesn’t eliminate third-party access risk. 

The Permission You Approved Three Years Ago 

Consider a common business scenario. 

A department buys software. IT connects it to Microsoft 365 and approves its permissions. It works, so everyone moves on. 

A year later, the department switches vendors. Maybe the original application is removed from everyone’s bookmarks. But did somebody revoke its cloud permissions? Did anyone check whether the old integration still has a credential or an administrative role? 

Repeat that for years of software purchases and trials. 

You can end up with dozens of trusted applications whose purpose is unclear, whose owners have left the company, or whose access is much broader than anyone remembers authorizing. 

An abandoned integration still creates a connection someone might abuse. 

The first step isn’t to rip out every third-party application. It’s to take inventory and put an owner next to each one. If nobody can explain why an integration is there, that’s a reasonable place to start asking questions. 

Trust Should Have Limits 

A common security principle is least privilege: give an account only the access it needs to perform its job. 

We apply this to employees all the time. Someone in accounting shouldn’t automatically have administrative control over every server. The same idea should extend to applications. 

If a scheduling tool needs calendar access, why should it read all company mail? If it needs a few SharePoint sites, why grant access to every site? 

Sometimes vendors legitimately require broad permissions. Backup and recovery tools are good examples because they need to protect data spread across an organization. But “the vendor requested it” shouldn’t be the end of the review. 

Ask what the permissions actually allow, whether narrower options exist, and what would happen if the application’s credentials were compromised. Some applications may need extensive access; leadership should at least understand the tradeoff. 

Review permissions when services are added, changed, or retired—not only after an incident makes the news. 

Would Your Security Team See the Warning Signs? 

Knowing which applications have access is one part of the problem. Knowing what they’re doing with it is another. 

Many organizations alert on suspicious employee logins and malware but have limited visibility into background application authentication. 

In the ConnectWise incident, researchers investigated unusual application sign-ins and changes to application credentials and privileged roles. Those aren’t the same signals as a user clicking a phishing link. They appear in different records, and they require someone to know what to look for. 

This is where continuous security monitoring matters. Your Security Operations Center, whether internal or provided by a partner, needs access to the relevant cloud identity and audit logs, not just alerts from employee laptops. 

Context matters. An overnight backup may be normal. An application suddenly signing in from unfamiliar infrastructure or assigning itself an administrative role isn’t. 

Collecting logs isn’t enough. Someone needs to investigate unusual activity and respond. 

A Vendor Can Fix Its Side and You Can Still Have Work to Do 

One of the more important lessons from this incident is that a vendor’s remediation doesn’t automatically resolve everything inside a customer’s environment. 

ConnectWise disabled affected connections and replaced credentials. By November 4, it reported backup and restore functions available for reauthorized Microsoft 365 tenants. But customers still needed to investigate their own environments and confirm that backups had resumed. 

An attacker who enters through one application may add credentials to another or change privileges, creating access that survives the original credential. 

Turning off the original connection doesn’t necessarily undo those changes. 

That means incident response needs to answer two different questions: Has the vendor stopped the initial problem? And has our organization checked whether anything was changed in our environment? 

Those aren’t the same investigation. In some cases, a qualified incident-response or digital forensics team is needed to establish what happened, preserve evidence, and determine whether access remains. 

Five Questions Worth Taking to Your IT Team 

Business owners, directors, and IT managers can start with five questions: 

  1. What outside applications can access our Microsoft 365 environment? Include old integrations, vendor management tools, and applications that work in the background. 
  1. Who is responsible for each one? Someone should know what it does, why the business needs it, and when it was last reviewed. 
  1. Which applications have the most powerful permissions? Prioritize anything that can access large amounts of data or make administrative changes. 
  1. Would we detect unusual application activity? Ask whether your monitoring covers application sign-ins, new credentials, permission changes, and unexpected administrative actions. 
  1. What would we do if a trusted vendor reported a compromise? Know who would investigate your own environment, decide what to disconnect, and verify that unauthorized access was removed. 

If the answers are unclear, that doesn’t automatically mean you’ve been compromised. It means you’ve identified a security gap that can be addressed before somebody takes advantage of it. 

What Better Vendor Access Management Looks Like 

Start with an inventory of integrations, an owner for each, a review of powerful permissions, and a process for removing access when a vendor relationship ends. 

From there, make sure your Microsoft 365 logs are being monitored for application activity and that your response plan covers compromised cloud identities, not just employee accounts and infected computers. Periodically test whether your team could actually investigate those events. 

A cybersecurity risk assessment can help identify unnecessary exposure. Ongoing SOC monitoring can help catch suspicious behavior before it goes unnoticed for weeks. And when an incident does occur, experienced responders can help determine its scope rather than assuming that the vendor’s fix solved everything. 

That’s the work we do at Alias Cybersecurity through risk assessments, SOC monitoring, incident response, and CISO support. We want businesses to use the tools they need while understanding the access and risk involved. 

The ConnectWise incident isn’t a reason to distrust every cloud provider. It’s a reminder that trust should come with visibility, limits, and a plan for what happens when something goes wrong. 

Your employees may be using MFA. Your third-party applications may already have the keys. Do you know who is holding them? 

If you’re unsure what has access to your Microsoft 365 environment, or whether your current monitoring would catch misuse, talk with Alias Cybersecurity. That’s a much better conversation to have before the next vendor advisory lands in your inbox. 

In this week’s episode of The SOC Brief, “When the Trusted App Becomes the Attacker — The ConnectWise Cloud Backup Incident,” we break down the attacker’s use of trusted application identities and the technical signs SOC teams should look for. 

Incident context reflects publicly available reporting through November 17, 2026. 

Written by: Andrew Hickman

Rate it

Previous post

Similar posts