ChannelBuzz.ca podcast

The Buzz: Attackers exploit ScreenConnect against MSP technicians, TD SYNNEX finds AI spending outpacing returns, and Microsoft details its outage

11/17/2026
0:00
7:07
Rewind 15 seconds
Fast Forward 15 seconds

If you run ScreenConnect, today’s top story needs your attention: Huntress says attackers are exploiting a critical flaw to take over MSP technicians’ workstations, and has seen a successful compromise. Meanwhile, a TD SYNNEX survey of 1,500 channel partners finds AI spending is running well ahead of the revenue it produces, and Microsoft’s preliminary report on last week’s outage points to a defect in its own networking gateway.

Those are the top stories in Tuesday’s edition of The Buzz from ChannelBuzz.ca.

Huntress confirms attackers are exploiting a ScreenConnect flaw to hit MSP technicians

Huntress says attackers are actively exploiting a critical ScreenConnect vulnerability against MSP technicians, and at least one attempt has already succeeded. This is a new development, separate from the SCKit extension attack covered in Monday’s edition, which Huntress ties to a different extension.

CVE-2026-84869 carries a CVSS score of 9.9. ConnectWise disclosed it on September 8, and the Canadian Centre for Cyber Security’s advisory says versions prior to 26.6.5 are affected. CISA added it to its Known Exploited Vulnerabilities catalogue on September 11. In research published Monday, Huntress explains that a guest in a remote support session can trigger file-transfer actions that the technician’s client runs without authorization or confirmation.

Huntress says it has seen at least 688,000 requests to run files across 42 MSP ScreenConnect servers from 29 IP addresses since October 20. On November 12, it confirmed a successful compromise: a batch file ran on a technician’s workstation and installed a rogue ScreenConnect client pointed at an attacker’s server in about nine seconds. On November 13, another technician’s client ran a harmless test file, which Huntress reads as the attacker confirming the exploit works.

Patching the server isn’t enough. Huntress says host clients must be updated to 26.6.5 or later, and it found technicians still running older 22.x, 24.2 and 25.8 clients weeks after the fix. If you can’t update right away, the interim advice, which Huntress attributes to ConnectWise, is to remove the TransferFiles permission from roles and session groups.

For MSPs, Huntress recommends checking server session data for guest-originated file events, looking for scripts run from ScreenConnect temp folders and unexpected Startup entries, and, if you find a hit, isolating the workstation and rotating every credential it could reach. Restrict who can join sessions as a guest.

TD SYNNEX survey: channel partners’ AI spending is outpacing the revenue it brings in

A new survey from TD SYNNEX suggests many channel partners are investing in AI faster than they’re earning from it. The distributor’s fifth annual Direction of Technology report, based on responses from 1,500 channel partners worldwide, was summarized by Channel Dive on November 11.

Two-thirds of respondents are accelerating AI investment, but just 21 per cent earn more than a quarter of their revenue from AI. At 45 points, that gap was the largest of any technology the survey studied. North American partners look further along than the global group: more than half are “AI ready,” meaning they’re actively investing, versus 44 per cent worldwide, and only 12 per cent are “AI aware” but not yet investing, versus 22 per cent globally.

The sticking point is getting from pilot to production. Only 13 per cent say most of their customers’ pilots or proofs of concept become scaled deployments. The top obstacles were security, legal and governance concerns, unclear return on investment, and a lack of internal skills.

There’s a services opportunity in that gap. Forty-two per cent of partners cite security, legal and governance concerns as deployment roadblocks, yet only 22 per cent sell governance services.

The broader business picture is healthier. More than three-quarters report revenue growth and 67 per cent report better margins, though 44 per cent point to pricing pressure and margin compression. Nearly half say existing relationships and trust help them win AI deals, and partners with mature internal AI programs report stronger revenue growth, margins and AI revenue.

For Canadian solution providers, the takeaway is that customers’ worries, not their enthusiasm, are shaping what they’ll pay for. Channel Dive’s summary doesn’t include a Canada-specific cut of the data.

Microsoft’s preliminary report blames a networking gateway defect for last week’s outage

Microsoft has published a preliminary post-incident report on last week’s Microsoft 365 outage, and it points to a software defect in its own networking gateway rather than an attack. The disruption began Thursday, November 12, and ran for nearly a full day, hitting Exchange Online hardest, plus Teams, Copilot, SharePoint and OneDrive.

These details come from a plain-language summary by 365Fanatic; Microsoft’s own wording governs. According to that summary, three factors combined. A latent defect in a fix for the gateway servers made tasks wait on each other under load, exhausting CPU and memory, which test labs missed on weaker hardware. A load surge followed when Teams automatically failed over to the gateway it shares with Exchange Online about 15 hours earlier. And nearly 14 per cent of the gateway’s capacity was already out of service because of an unrelated issue the day before. A separate problem then stopped normal isolation between components and regions, so the failure cascaded globally.

Microsoft’s earlier public explanation was less specific. It said the problem was a technical issue and not a cyberattack, tracing it to a network capacity upgrade that surfaced after a recent network change, as Forbes reported. The preliminary report frames the cause differently, so watch the final version for a reconciled account.

Microsoft began reverting the defective code path around 07:20 UTC on November 13, returned servers to service in stages, added capacity, and extended the email delivery window so delayed messages weren’t bounced. A final report with a full list of repair items is expected November 20, per the summary.

For MSPs, the report gives you a sourced account to share with customers who asked what went wrong. It’s also a prompt to check what your clients’ continuity plans assumed about Exchange Online.

In The Channel: Yesterday on In The Channel, D&H Canada builds Advanced Solutions+ by teaching resellers to fish, recorded at the distributor’s THREAD conference.

Later this morning at nine, ACT360’s Adam Bowles on putting people at the centre of cybersecurity: why rising security spending hasn’t stopped rising breaches, and what MSPs can do about it.

Read Full Transcript

Welcome to The Buzz for Tuesday, November seventeenth, twenty twenty-six. I’m Robert Dutt, and here’s what’s happening in the channel today.

If you run ScreenConnect, you’ll want to hear today’s top story. Huntress says attackers are exploiting a critical flaw to take over MSP technicians’ workstations, and it has confirmed a successful compromise. Plus, a TD SYNNEX survey of fifteen hundred channel partners finds AI spending running well ahead of the revenue it brings in, and Microsoft’s preliminary report on last week’s outage points to a defect in its own networking gateway.

Up first, and if you use ScreenConnect, this one needs your attention today.

Huntress says attackers are actively exploiting a critical ScreenConnect vulnerability against MSP technicians, and at least one attempt has already worked. This is a new development, separate from the SCKit extension attack we covered on Monday. Huntress ties that one to a different extension.

The flaw is tracked as CVE twenty twenty-six, eighty-four eight-six-nine, and it carries a severity score of nine point nine. ConnectWise disclosed it on September eighth, and the Canadian Centre for Cyber Security says versions before twenty-six point six point five are affected. CISA added it to its Known Exploited Vulnerabilities list on September eleventh. Huntress explains that a guest in a remote support session can trigger file-transfer actions that the technician’s client runs without asking for authorization or confirmation.

Here’s what Huntress is seeing. Since October twentieth, it has counted at least six hundred eighty-eight thousand requests to run files, across forty-two MSP ScreenConnect servers, from twenty-nine IP addresses. On November twelfth, it confirmed a real compromise. A batch file ran on a technician’s workstation and installed a rogue ScreenConnect client, pointed at the attacker’s server, in about nine seconds. The next day, another technician’s client ran a harmless test file, which Huntress reads as the attacker checking that the exploit works.

And patching the server isn’t enough. Huntress says the technician’s client has to be updated to twenty-six point six point five or later, and it found technicians still running older clients weeks after the fix came out. If you can’t update right away, the interim advice, which Huntress attributes to ConnectWise, is to remove the TransferFiles permission from roles and session groups.

So what should you do? Update every technician client. Check your server session data for file events started by guests. Look for scripts running out of ScreenConnect temp folders and for Startup entries you didn’t create. And if you do find a hit, isolate that workstation and rotate every credential it could reach. It’s also worth restricting who can join your sessions as a guest.

Next, a new survey suggests a lot of channel partners are spending on AI faster than they’re earning from it.

TD SYNNEX’s fifth annual Direction of Technology report is based on responses from fifteen hundred channel partners worldwide. Channel Dive summarized it on November eleventh.

Two-thirds of respondents are accelerating their AI investment, but only twenty-one per cent earn more than a quarter of their revenue from AI. That forty-five point gap was the largest of any technology the survey looked at. North American partners look further along than the global group. More than half call themselves AI ready, meaning they’re actively investing, compared with forty-four per cent worldwide. And just twelve per cent are AI aware but not yet investing, compared with twenty-two per cent globally.

The sticking point is getting from pilot to production. Only thirteen per cent say most of their customers’ pilots become scaled deployments. The top obstacles are security, legal and governance concerns, unclear return on investment, and a lack of skills in-house.

And there’s a services opportunity hiding in there. Forty-two per cent of partners cite security, legal and governance concerns as roadblocks, yet only twenty-two per cent sell governance services.

The broader picture is healthier. More than three-quarters report revenue growth, and sixty-seven per cent report better margins, though forty-four per cent point to pricing pressure. Nearly half say existing relationships and trust help them win AI deals.

The takeaway to think about is that your customers’ worries are shaping what they’ll pay for.

And finally, Microsoft has published a preliminary report on last week’s Microsoft 365 outage, and it points to a software defect in its own networking gateway, not an attack.

The outage began Thursday, November twelfth, and ran for nearly a full day. Exchange Online was hit hardest, along with Teams, Copilot, SharePoint and OneDrive.

The details here come from a plain-language summary of the report by 365Fanatic, so Microsoft’s own wording governs. According to that summary, three things combined. A hidden defect in a fix for the gateway servers made tasks wait on each other under load, using up CPU and memory, and test labs missed it because they ran on weaker hardware. Then Teams automatically failed over to the gateway it shares with Exchange Online about fifteen hours earlier, which brought a surge of load. And nearly fourteen per cent of that gateway’s capacity was already out of service because of an unrelated issue the day before. A separate problem then stopped normal isolation between components and regions, so the failure spread globally.

Microsoft’s earlier explanation was less specific. It said this was a technical issue, not a cyberattack, and traced it to a network capacity upgrade that surfaced after a recent network change, as Forbes reported. The preliminary report frames the cause differently, so watch the final version for a reconciled account.

Microsoft began reverting the defective code around seven twenty UTC on November thirteenth, returned servers to service in stages, added capacity, and extended the email delivery window so delayed messages weren’t bounced. A final report with a full list of repair items is expected November twentieth.

For MSPs, this gives you a sourced account to share with customers who asked what went wrong. It’s also a good prompt to check what your clients’ continuity plans assumed about Exchange Online.

Before you go, if you missed it, D&H Canada’s Dan Reio and Colin Blair were on In The Channel yesterday, talking about how the distributor builds its Advanced Solutions Plus business by teaching resellers to fish. And later this morning at nine, Adam Bowles, the CEO of Barrie, Ontario MSP ACT360, joins us to talk about putting people at the centre of cybersecurity.

That’s how we’re seeing the headlines today. I’m Robert Dutt. Thanks for listening. Have a great day.

More episodes from "ChannelBuzz.ca"