All posts

The Attacker Brought Your Kind of IT Tool

A

Antonio Pagano

Author

7 min read
Share:
The Attacker Brought Your Kind of IT Tool

TLDR: In the last week of October, Huntress described phishing that got users to enroll their own Windows PCs in device management (MDM) tenants the attacker controlled, then used that trusted path to install ScreenConnect.[1] In one case the MSP removed ScreenConnect and it came back hours later through the MDM client that was still there. In the other, a fake Teams invite enrolled a receptionist’s PC in about a minute. Two days later, Cisco Talos described fake IT help desk callers who got by on little more than names and email addresses.[2] Every tool in these stories is legitimate. For MSSPs, the habit to teach users is small: an IT request you did not expect is the moment to verify through a channel you started, and to report. The number to watch is how fast those reports arrive.

Every MSP I know teaches clients the same thing in onboarding, usually without saying it out loud: IT installs things on your computer. An agent shows up. A management profile appears. Someone from support asks you to open a session so they can take a look. Users learn that this is normal, because it is.

Attackers learned it too.

A real installer, the wrong tenant

Huntress published two incidents on October 28.[1]

In the first, a user received emails with proposal-style subjects like “ATTN: Official Invitation to Submit Proposal” that mentioned an organization the user likely worked with. The link downloaded a file named eD0cuSign_Installer.msi. It was not DocuSign. It was a legitimate Hexnode MDM package, and it installed Hexnode’s gateway and agent. The attacker then installed ScreenConnect, which connected to an attacker relay. Huntress alerted, and the MSP removed ScreenConnect.

Several hours later, ScreenConnect was installed again and reconnected to the same relay. The MDM client the attacker had enrolled was still on the machine, and it simply delivered the tool a second time. Huntress isolated the endpoint and remediation started.

In the second, a receptionist opened an email titled “Re: Tax Planning and Preparation Inquiry” with what looked like a Microsoft Teams meeting link. It led to a page titled “Calendar” on a lookalike Teams domain, which helped enroll the PC in an attacker-controlled Rippling MDM tenant. About a minute after the click, the enrollment showed up on the system. The Rippling agent installed with SYSTEM permissions. About 35 minutes later, the attacker used the MDM to run a ScreenConnect installer as SYSTEM.[1]

About a minute. Most legitimate device enrollments I have watched take longer than that, and those are the ones we planned.

Huntress is clear that neither Hexnode nor Rippling is at fault. Users were socially engineered into installing real software.[1] That is the part that matters for an awareness program.

The caller who only needed a name

On October 30, Help Net Security covered a Cisco Talos report on fake IT help desk callers.[2] The callers reached people by phone, email and Microsoft Teams, and they needed little more than names and email addresses to sound credible. They did not go only after admins. They went after whoever would say yes. In one case, the victim called a phone number shown in a scam pop-up. In another, a vishing call led to a Quick Assist remote session, and the intrusion ended in ransomware.[2]

Talos also reported that phishing was still the most common initial access method in its engagements that quarter, at 37%.[2]

Put the two reports side by side and the shape is the same. The attacker does not need the user to run something that looks dangerous. The attacker needs the user to accept an IT request that looks routine: install this, join this meeting, let me connect to your screen.

Removing the tool is not the end

The Hexnode case has a detail every MSSP should keep. The visible part of the intrusion was ScreenConnect. The MSP removed it, correctly. It came back because the thing that delivered it was still enrolled.

Huntress puts it this way: removing the remote tool does not remove the unauthorized management path that delivered it, and MDM enrollment should be treated as security-sensitive activity, not a routine admin task.[1] Their mitigation list starts with keeping an inventory of authorized MDM products, tenants, agents and RMM instances, and treating any unexpected enrollment, management agent or RMM instance as a security event.[1]

That is the technical side. It is also an awareness point, because the fastest way to learn that a PC was enrolled in the wrong tenant is a user who says, “I installed something for a meeting and now I have a new icon.”

What to teach users

Huntress recommends training users to verify unexpected meeting, document-signing and proposal-review requests.[1] Talos recommends help desk verification before new MFA enrollment.[2] Both point at the same habit, and it is smaller than most training makes it.

The trigger is the request, not the file. Users cannot tell a real Hexnode package from a malicious one, and they should not have to. They can notice that a meeting invite asked them to install something, that a signing request came with an installer, or that “IT” called when they had not asked for help.

Verify through a channel you started. Call the help desk number you already have. Message your IT contact in the tool you already use. Do not use the number in the pop-up, the link in the invite, or the callback the caller offers.

Report even after you said yes. The receptionist in the Rippling case had about a minute before enrollment and about 35 more before ScreenConnect arrived.[1] A report inside that window changes the story. A user who is afraid of being blamed for clicking will wait. A user who knows reporting is the job will not.

What to measure

This is why I care more about report rate than click rate. A click rate tells you how many people would have started the chain. A report rate, and how quickly reports arrive, tells you how many would have stopped it in the first hour.

For MSSPs running programs across many clients, three practical moves follow:

  1. Run simulations that ask for an action, not just a click. Lures framed as a meeting invite that needs a quick install, or a document that needs a viewer, rehearse the exact request in these reports.
  2. Tell users what your real IT requests look like. If your clients know how your team actually asks them to install an agent or open a remote session, an unexpected request stands out.
  3. Connect the report to the inventory. When a user reports an install they did not expect, the analyst’s first question should be “what is enrolled on this device, and is it ours,” not only “is the file malicious.”

Where Symbol fits

Symbol helps admins run security awareness across many tenants without a person for every client. Our phishing simulation library is where users rehearse the moment an ordinary-looking request asks for more than a click. Security awareness training is where the verify-through-your-own-channel habit gets taught across every client at once. And Report a Phish gives users a single action for “this IT request feels off,” so the report reaches your analysts while the window is still open.

If you want to see how your own employees respond when the lure looks like routine IT, try the program on them first. That is usually the most honest test.


References

  1. Huntress, “Managed for Malice: How Attackers Turn MDM into Remote Access,” October 28, 2026. https://www.huntress.com/blog/attackers-turn-mdm-into-remote-access
  2. Help Net Security, “Fake IT help desk callers got by on little more than names and email addresses,” October 30, 2026, reporting on Cisco Talos incident response findings. https://www.helpnetsecurity.com/2026/10/30/cisco-talos-it-help-desk-vishing/
A

Written by Antonio Pagano

Antonio Pagano is Co-Founder and CTO of Symbol Security.