Home/Blog/ScreenConnect MSP attacks: what your MSP must check
blog

ScreenConnect MSP attacks: what your MSP must check

November 12, 2026 · Security News

ScreenConnect MSP attacks: rogue remote-access clients pushed from compromised provider consoles onto customer PCs.

Huntress is tracking ScreenConnect MSP attacks that pushed rogue remote-access clients from compromised provider accounts onto customer PCs, with six MSPs and 47 organizations in the first count.

Key takeaways

  • Huntress tracked six compromised MSPs, 286 or more endpoints, and 47 organizations, and said the counts were still climbing.
  • The campaign abused ScreenConnect Run Command after partner access was already in hand; no CVE has been named.
  • ConnectWise disabled five extensions across about 400 cloud partner environments; on-premises customers must disable them themselves, and disabling does not remove rogue clients or persistence.
  • Technicians should not open interactive ScreenConnect sessions to potentially compromised devices while workstation-compromise reports are unconfirmed.

What happened in the ScreenConnect MSP attacks?

Huntress is tracking an active campaign in which attackers used compromised ScreenConnect environments at managed service providers to install rogue remote-access clients on customer PCs. ConnectWise is investigating. For a business that pays an MSP, the issue is a second remote-control channel beside the one the provider already uses.

Huntress said its SOC found the activity on Wednesday and that the campaign started around 17:47 UTC that day. In a Reddit post quoted by BleepingComputer, the SOC wrote that it was tracking six compromised MSPs and 286 or more endpoints, with those numbers still climbing. The endpoints sat across 47 organizations.

Spike Brandt, principal threat intelligence incident commander at Huntress, told BleepingComputer the activity appears to be abuse of legitimate ScreenConnect functionality. Huntress has not named a CVE. The instances it described were cloud-hosted.

The feature under abuse is Run Command, which executes commands on managed endpoints in the agent’s own context. An operator who already has a valid session can push an installer the same way a technician pushes a legitimate fix.

The rogue software was another ScreenConnect client, so the attacker could reach the same PC from an environment the MSP does not control. That second client is the artifact to find on the machines your MSP already manages.

Initial access is still unconfirmed. Huntress told BleepingComputer that current telemetry points to compromised credentials or hijacked operator sessions, and Brandt said the team has no information about how the first compromised machine or account was breached.

Huntress began receiving alerts when rogue clients appeared on endpoints it already monitors, then retrieved the attacker’s scripts, which were still downloadable from the staging URLs. Sequential attacker-assigned tags, Huntress told BleepingComputer, suggest other MSPs may have been hit earlier. Review Run Command history even when no alert fired.

Which ScreenConnect extensions did ConnectWise disable?

ConnectWise disabled five third-party extensions across about 400 cloud-hosted partner environments: Labtech Plus, RMM Plus, RMM+, RMM Plus Password Link, and RMM+ Password Link. The vendor published that step Wednesday night, after confirming unauthorized access to ScreenConnect partner accounts and unauthorized remote-access software in customer environments. Partners were told not to re-enable or reinstall the five extensions until further guidance.

ConnectWise’s initial assessment is that the threat actor deployed additional remote-access software through independent third-party extensions partners may have added. Having those extensions installed does not, by itself, mean a network was compromised. ConnectWise said it has no reason to believe partners that have not deployed the five identified extensions are impacted, but it still asked every partner to stay on heightened alert.

Disabling the extensions does not remove malware, persistence, or access already on devices. ConnectWise said the disable action does not establish that unauthorized software, persistence mechanisms, or access already present has been removed. Treat a disabled add-on as containment only.

ConnectWise turned the five extensions off in the cloud instances it hosts. On-premises ScreenConnect customers have to disable them manually, because ConnectWise cannot do that from its side. If your provider runs ScreenConnect on its own servers, ask for confirmation that the five names are absent or disabled, and ask what was checked on the PCs themselves. A missing extension is not a clean bill for Run Command history or for the PCs.

What should an SMB ask its MSP today?

Ask three things in one note: whether ScreenConnect is cloud-hosted or on-premises, whether any of the five named extensions were ever installed, and whether a second ScreenConnect client is on your PCs. ConnectWise told partners to inspect authentication logs, extension installation records, administrative activity, Run Command history, and Toolbox records.

  • Install time and installing account for Labtech Plus, RMM Plus, RMM+, RMM Plus Password Link, and RMM+ Password Link, or a written statement that none were present.
  • Run Command entries that downloaded or launched an MSI or an EXE, especially another ScreenConnect installer or a script from an external address.
  • New ScreenConnect client services or processes, and any client identity that is not in the provider’s inventory.
  • Authentication events on and before the Wednesday Huntress named, starting from around 17:47 UTC, looking for unusual source addresses, odd hours, and successful logins.

Huntress published indicators, including rogue ScreenConnect client IDs, attacker infrastructure, a stager URL, an installer pattern, an attacker-assigned name, and tool filenames. Take those indicators from Huntress and match them in your own tools, or ask the MSP to run that match and send the result. A copied list goes stale once the infrastructure moves.

Valid accounts are the suspected path, so a quiet failed-login report is a weak signal. After RMM abuse, incident responders look for a successful admin action at a strange hour, a Run Command that fetches an installer, and a new remote-access service that appeared the same day.

If the ScreenConnect login or the partner mailbox was taken over, account takeover recovery is the work of putting the real operators back in control and listing what the intruder opened during the session.

Ask the MSP, in writing, whether anything on your tenant is still changing. Huntress said its counts were climbing when it posted. It also contacted ConnectWise, the VPS provider hosting the attacker’s infrastructure, cloud providers whose dynamic DNS services were used, and the broader MSP community.

If a rogue client is present, isolate that PC and check the rest of the fleet. Threat hunting is how you learn whether the intruder is still on the network after the obvious installer is gone. A single unexpected remote-access install is a strong indicator.

Should technicians avoid interactive ScreenConnect sessions?

Yes. ConnectWise told technicians not to start interactive ScreenConnect sessions with potentially compromised devices, and to investigate with separate security tools instead. The instruction protects the technician laptop as well as the customer PC.

ConnectWise has received reports of technician workstations compromised during interactive sessions and said those reports are still being confirmed. The mechanism is not public. Until the vendors describe how that jump works, stay out of the interactive session and collect evidence with another tool. If engineers already opened an interactive session after the rogue client landed, treat those workstations as in scope.

Custom tooling was sometimes compiled on the victim computer, Brandt told BleepingComputer, so a hash from yesterday may not match the file on disk today. Check recent logons, new remote-access software, and scheduled tasks on the technician PC the same way you check the customer endpoint.

A managed detection and response service that watches the customer fleet and the MSP’s own admin machines makes that second check routine. Huntress caught the rogue clients because the endpoints were already monitored.

How does the July 2026 script fit, if it fits at all?

One July 2026 batch script shares a rogue ScreenConnect client ID with the Huntress list, and nobody has shown it belongs to this campaign. BleepingComputer found the file, first uploaded to VirusTotal on July 25, 2026. Huntress declined to comment on that specific July incident.

The July behaviors are still searchable. The script impersonated a Social Security Administration statement, added Microsoft Defender exclusions, and set persistence through the Windows Startup folder and a Run registry key. It contacted an external domain for installation commands covering two separate installers, and VirusTotal sandbox activity showed downloads of ScreenConnect MSI installers. The script then used PowerShell to search running processes and Windows services for a rogue client ID later listed by Huntress. If that ID was present, the script sent a Telegram notification.

The Social Security filename points to a phishing lure, a different distribution path from this week’s abuse of an MSP console. Huntress said it has seen a significant increase in 2026 in attacks that use legitimate RMM software as an initial payload, but the link between the July file and the current attacks remains unclear. Use the July file as a lead only when you already have the Huntress IDs.

How do you check for a rogue ScreenConnect client?

A rogue ScreenConnect client is a second remote-support agent, installed through the provider’s own Run Command, that reports to an environment the MSP does not run. It can sit beside the legitimate agent. Seeing the usual icon is not a clean result.

On Windows, look for a ScreenConnect service or process whose client identity is absent from the provider’s inventory. Compare that install time with Run Command history. An MSI launched from the agent, with an installer argument the MSP does not recognize, is the event to pull first. Also look for what persists after an installer finishes: a Defender exclusion, a Startup-folder item, a Run key, and a second remote-access product that is not ScreenConnect.

Cloud customers still need the endpoint check. ConnectWise can disable an extension on a hosted instance. It cannot see a service that already landed on a laptop in your office. On-premises customers do both jobs: turn the five extensions off in the admin interface, then hunt the endpoints.

What is still unknown?

Nobody has named the attackers. Nobody has confirmed how the first machine or account was breached. Nobody has published how an infected endpoint could compromise a technician workstation. What you can settle today does not wait on attribution: confirm whether the five extensions were present, whether Run Command history shows an installer, and whether every ScreenConnect client on your PCs belongs to your MSP.

If the extensions or an unexpected Run Command entry turn up, or if the provider cannot produce the logs, treat the machines as carrying unauthorized remote access and start data breach response the same day. A threat-actor name will not uninstall a client that is already listening.

Same-day checks for an owner:

  • Ask whether the instance is cloud or on-premises, and whether the five extensions are disabled.
  • Ask for Run Command history covering installer launches, including commands run before any alert.
  • Ask which ScreenConnect client identities are approved, and which ones were found on your PCs.
  • If a technician connected interactively after a rogue client appeared, include that laptop in the same review.

FAQ

What is the ConnectWise ScreenConnect attack?

Attackers used compromised ScreenConnect environments at MSPs to install separate, unauthorized ScreenConnect clients on customer machines. Huntress found the activity on Wednesday, starting around 17:47 UTC, focused on cloud-hosted instances.

Which ScreenConnect extensions did ConnectWise disable?

Labtech Plus, RMM Plus, RMM+, RMM Plus Password Link, and RMM+ Password Link, across about 400 cloud-hosted partner environments. On-premises customers must disable the same five themselves.

Is this ScreenConnect campaign a software vulnerability?

Huntress says it looks like abuse of legitimate ScreenConnect features, including Run Command. No CVE has been named for this campaign.

Does disabling ScreenConnect extensions remove the malware?

No. ConnectWise said the disable action does not show that unauthorized software, persistence, or access already on devices has been removed. Endpoints still need a separate check for rogue clients.

If a second remote-access client is already on your PCs, our data breach response team can separate your provider’s tools from the intruder’s and help you take the machines back.


← All articles

get in touch

Tell us what you're running

Scoping is free. We reply within one business day, and under 30 minutes for active incidents.