If your IT provider manages your computers with ConnectWise ScreenConnect, or backs up your Microsoft 365 data with ConnectWise Cloud Backup (formerly SkyKick), you should ask them some questions this week. ConnectWise posted updates on both incidents between 13 and 18 November. Each update leaves work for the managed service provider (MSP) to do, and some of that work affects the provider's customers.
What happened inside ScreenConnect
On 11 November, security firm Huntress reported rogue ScreenConnect clients on at least 288 managed endpoints across 48 organisations. All of those machines were hit within about 90 minutes. The attackers used ScreenConnect's own Run Command feature to install a second remote access client along with a toolkit Huntress calls SCKit. Huntress then logged a second wave on 13 November that reached another 205 hosts across 41 organisations from new infrastructure.
Huntress thinks the attackers got in through RMM+, a third-party ScreenConnect extension. In versions 1.0.31 and 1.0.32, if an administrator never set the extension's access key, RMM+ accepted a blank key. The extension's API runs commands without going through ScreenConnect's user accounts, roles or multi-factor authentication. ConnectWise disabled five related extensions across about 400 partner environments and said the initial access method was still unconfirmed. ConnectWise also said that having one of those extensions installed didn't mean an environment had been compromised.
SCKit's design makes the incident worse for anyone who responds to it. One component waits for a technician to connect to an infected machine, then tries to push a malicious file back to the technician's workstation and steal ScreenConnect session cookies. With those, the attacker could reach every other customer the technician manages. According to BleepingComputer, the component targets CVE-2026-84869, which ConnectWise fixed in version 26.6.5. Blackpoint couldn't reproduce the attack against a current client, and Huntress found the console takeover had failed in every recovered log it examined.
ConnectWise's Trust Center advisory added more on 13 November. Partners were told to disable the RMM+ family of extensions and to confirm technician clients were running 26.6.5 or later. In its cloud environments and marketplace, ConnectWise disabled three more integrations: ITBoost, RepairShopr and Guest Session Starter. On 16 November ConnectWise said its investigation was complete. In the same update it said researchers were still seeing activity, that ConnectWise can't disable extensions on partners' on-premises servers, and that closing the investigation didn't show any individual environment was unaffected.
The backup service is back, but each tenant needs action
The second incident started in October. On 16 October, ConnectWise found that someone was using Cloud Backup's application credentials without authorisation, and it shut the service's access off the next day. On 21 October ConnectWise linked the activity to a China state-aligned group, citing Microsoft threat intelligence. A summary of the public reporting says Microsoft counted more than 2,250 Microsoft 365 tenants targeted. That number counts attempted sign-ins, and it doesn't measure confirmed breaches.
ConnectWise's 18 November update says backups, restores, monitoring, reporting and exports are working again. The service now runs on a new application that uses federated identity credentials, so there are no stored secrets for an attacker to steal. Each Microsoft 365 tenant still has to be re-authorised by the partner, and ConnectWise is blunt about it: "Partner action is still required to resume protection." Re-authorising a tenant doesn't prove backups are running, so ConnectWise tells partners to check that new snapshots actually complete. SaaS Security and Cloud Automation customers also have to reconnect their Microsoft integrations.
If a tenant hasn't been re-authorised, its mail, OneDrive and SharePoint data may have had no new backups since 17 October. The business owner probably has no idea. Status pages and green dashboards don't show it, and the provider may have hundreds of tenants to work through.
Why NZ businesses should care
Both incidents went through tools the MSP trusts, with access the customer approved long ago. An unset extension key bypassed MFA on a remote support console. Vendor application credentials gave access to Microsoft 365 tenants. In each case the customer's own controls didn't matter much, because the attacker came in through the provider's management layer.
Most small and mid-sized NZ organisations hand their device management and backups to an MSP, and their service agreements rarely say how the provider manages its own tools. Supplier management, configuration management and change control for third-party add-ons are standard ITSM practices, and an ITSM practice like the one in our ITIL 4 framework work should apply them to whatever sits between your provider and your machines. A backup that nobody has tested after a vendor incident counts as an untested assumption, and it shouldn't be treated as a working control.
Some facts are still unpublished: how the attackers first got the Cloud Backup credentials, the confirmed initial access path into ScreenConnect, and how many customer organisations had data accessed. Plan as if your environment could be one of them until your provider shows evidence that it isn't.
Next steps
- Ask your MSP in writing whether they use ScreenConnect, whether any RMM+ family extension was ever installed, and whether every technician client now runs 26.6.5 or later.
- Ask whether they've reviewed Run Command and Toolbox history for your endpoints and checked for a second, unauthorised ScreenConnect client. The SCKit indicators Huntress published include the TrapWatch scheduled task and the
C:\ProgramData\RelayTrap\folder. - If you're on ConnectWise Cloud Backup, ask for the date your tenant was re-authorised and the timestamp of the latest successful snapshot. Then ask for a test restore of one mailbox and one SharePoint site.
- Find out whether your Microsoft 365 data went without backups between 17 October and re-authorisation, and decide whether you need a one-off export to cover that period.
- Check Entra ID for enterprise applications and consents tied to old ConnectWise or SkyKick apps, and remove any your provider no longer needs.
- Add a clause to your MSP agreement that requires notice within an agreed time when a tool with access to your systems has a security incident, along with a written list of the third-party extensions they run.
- If your own team runs ScreenConnect on-premises, disable the affected extensions yourself, because ConnectWise can't do it for you. Investigate suspect machines with separate tooling and don't open interactive sessions to them.
If your providers can't answer these questions quickly, take that as a sign your supplier controls need work. Our process and automation team can help turn checks like these into a regular service review.
Sources
- SCKit Worm Installs Rogue ScreenConnect Clients and Session Cookie Relays Across Hundreds of Endpoints (Huntress)
- ConnectWise Trust Center: Security advisories
- ConnectWise disables third-party extensions after cyberattack on ScreenConnect (Cybersecurity Dive)
- New worm-like SCKit malware targets MSPs in ScreenConnect attacks (BleepingComputer)
- ConnectWise investigating reports of unauthorized access to third-party extensions (SC Media)
- ConnectWise Cloud Backup breach: 2,250 tenants (shattered.io)
Want to know what this means for your organisation? Our team in Auckland can show you your top three opportunities in a free 30 minute assessment.
Get my free assessment How we help