Opens in a new tab
News, Trends, and Insights for IT & Managed Services Providers
News, Trends, and Insights for IT & Managed Services Providers
6fa8c3fd e835 4129 ad4a e447d00b9b03

Patched, and Still Breached
Before we start, disclosures: I own stock in N-able, and N-able is part of today’s story.  Huntress supports this episode.

The patch used to be where a security problem ended. The evidence says it’s now where one starts.

Start with Google. The Register reports that Google’s Threat Intelligence Group found an AI framework called exploit_agent on an exposed server, tied to a suspected Russian-speaking operator. It was set up to keep comparing popular web application software before and after security updates. In plain terms, it reverse-engineered each patch into an attack, with no human step. The same server held data stolen from victims, plus scripts for brute-forcing passwords and planting web shells. Google calls it the first case it has seen of an AI agent turning patches into working attacks with no human in the loop. That’s Google’s own threat research, and the report doesn’t say how often those exploits succeeded.

Now bring it closer to home. The Hacker News reports that attackers installed cryptocurrency miners on 21 N-central servers, according to security firm Huntback. Nineteen of those 21 had already applied Hotfix 2, N-able’s previous fix. Huntback’s advice: “Treat a hit as a full compromise, not a cleanup job.” N-able has since shipped Hotfix 3, covering 17 more vulnerabilities, and removed the legacy API those attacks depended on. N-able says the attackers were going after customers who still had that legacy API turned on. It also says every customer, hosted or on-premises, has to update its agents and probes, because ones that aren’t upgraded “are susceptible to impersonation.”

Then look at ConnectWise. It says attacks on its ScreenConnect remote-access tool are continuing against on-premises systems where certain third-party extensions are still enabled and the software hasn’t been updated. It disabled five of those extensions across 400 partner environments, while noting that having one installed doesn’t by itself mean a compromise. And Huntress saw another spike in activity from the attackers’ toolkit on November 13, across 205 hosts at 41 organizations.

Hear what that adds up to. An attacker’s AI can read a published fix and build an attack from it on its own. Two remote management platforms MSPs run their businesses on are under active attack. And on one of them, almost every server that got hit had already installed the fix.

In plain terms, “we patched it” no longer tells you whether you’re safe.

So why doesn’t a fix end the story anymore?

What a Patch Can’t Fix
Underneath all of this is one dynamic. A patch changes code, and only code. It closes the hole it was written for. It does nothing about whoever already came through that hole, and it can carry a new failure of its own. That used to be manageable, because attackers needed time to work out what a patch fixed, and defenders used that time to install it. AI is taking that time away, and what’s left exposed is everything a patch was never built to cover.

Start with what a patch leaves behind. The Record reports that Blackpoint Cyber reverse-engineered the toolkit used in the ScreenConnect attacks. The existing patch blocks the vulnerability it exploits. But the toolkit can also hijack a technician’s logged-in session and push malicious code through ScreenConnect’s legitimate features. Blackpoint’s conclusion: “Patching addresses the known exploit, but it doesn’t eliminate the risks associated with a compromised operator session.” ConnectWise warns that disabling extensions and resetting credentials may not remove software attackers already installed.

Now look at what a patch can bring in. Microsoft executive Charles Lamanna has explained the November 12 Microsoft 365 outage, and iTnews reports the bug came in through a fix Microsoft wrote for its own networking gateway servers. That fix rolled out from July and ran worldwide for more than three months before it failed. Microsoft’s load tests ran on weaker hardware with fewer processor cores, where the problem didn’t show. When it hit, more than 80 percent of the gateway’s servers stopped responding in under two hours. In Lamanna’s words, “our normal rollback, flighting, and feature flags were unable to quickly revert the change.”

Now, a fair objection: attackers have studied patches for years. That’s true. Comparing code before and after a fix to see what changed is an old technique. What’s new, as The New Stack puts it, is that agents can now do much of that work themselves. After two PaperCut flaws were disclosed in August, security researchers and Huntress built working exploits with AI help within a day. Four days after disclosure, GreyNoise documented an attacker who likely used AI to build exploits and target lists, then went from first access to domain administrator within minutes. Everything the patch didn’t cover becomes the place attackers work.

The federal government has already rebuilt around that. Federal News Network reports that FedRAMP is requiring more than 550 certified cloud providers to move off monthly scans and severity scores. In their place comes an approach built on exposure and threat: can the flaw be reached from the internet, and how exploitable is it?

A patch is like changing the locks. It keeps the next person out. It doesn’t tell you who’s already in the house, and sometimes the new lock jams. Those used to be edge cases.  Now they’re the job.

That job shows up in front of the client.

The Job After the Patch
For the MSP, the consequence is that patching is now the floor of the service, not the service. What clients actually need is what happens around the patch: knowing nothing got in before it, and keeping the business running when the patched system doesn’t.

Start with how organizations rode out the Microsoft outage. Advisory firm Infotech, which sells continuity-planning research, found that contingency plans held “where a manual route existed” and stalled “where email was the only channel.” It told CIOs to confirm that critical workflows, integrations and security controls had actually recovered before telling the business it was back to normal. It also warned that attackers often pose as support staff during outages. That’s the job after the fix: not “the service is up,” but “the business works, and nothing slipped in.”

Then there’s the other end: systems where no patch is coming at all. The Register reports that Microsoft now says Exchange Server 2016 and 2019 “are no longer tested for vulnerabilities and there are no updates released.” The final security updates shipped in October. Microsoft’s November fixes for its current Exchange version included a critical remote code execution flaw, and it isn’t testing the older versions for it. For a client still running one of those servers, a patch schedule has nothing left to offer. Microsoft’s advice: “upgrade your organization immediately.”

Put those together. Whether the fix arrives, fails, or never comes, the client’s safety depends on what the MSP does beyond installing it. There’s some good news in Google’s report. These AI attack tools are noisy. They leave heavy network traffic and recognizable code behind, and Google says attackers currently “trade stealth for velocity.” The MSP who goes looking after a patch has a real chance of finding them.

So here’s the choice. Report what you verified. After every critical patch on your own tools and on anything a client exposes to the internet, check for anything that got in, reset technician sessions, confirm the work still runs, and give any process that depends on a single system a second path. Or keep reporting that you patched, and let a green compliance dashboard answer a question nobody checked.

That choice starts with a conversation most clients have never had.

Why Do We Care?
Clients have been told for years that patched means protected, and the next breach or outage will arrive after a green report. Before it does, tell each client in plain words what a patch covers and what it doesn’t: it closes the hole, it doesn’t prove nobody got in, and it doesn’t keep their email running. Then, after the next critical patch, show them what you checked.

What to Consider

  • Rewrite the patch report. Replace the compliance percentage with three lines for each critical patch: what was patched, what you checked afterward, and what you found. Clients learn what “patched” means from what you put in front of them, so change what you put in front of them.
  • Name the processes that only run on email. Ask each client which approvals, payments and customer notices travel only by email, because processes with no route but email are what stalled during the Microsoft outage. Give each one a documented second path, and tell the client which path it is before they need it.
  • Tell clients which patches aren’t coming. List every client system that’s past end of support and reachable from the internet, starting with Exchange 2016 and 2019. Put it in writing that no fix is coming, with a date to migrate it or take it off the internet.

If this trend continues: By the end of 2027, the gap between a published patch and a working exploit will be hours for widely used software, and after an incident a client’s first question will be what you checked, not whether you patched.

This content is made possible by community-minded viewers like you

Join the Small Biz Thoughts Community to support independent journalism and get the raw Business of Tech interviews, Channel Intelligence briefs, monthly member calls and more. Member support is one way we sustain the costs of investigative work — from court filings to video editing — while keeping the core reporting open to all.

Thank you

Join the Small Biz Thoughts Community

Choose your upgrade:

Get the full benefits of Business of Tech Plus

Insider Access

$12/month

Perfect for MSPs and ITSPs that want full interviews, early access, and ad-free listening

  • Programmatic Ad-free private podcast feedSame show, little interruptions
  • Channel Chatter previews1–2 topics with light insights
  • Early access to interview episodesHear it days before public release
  • Monthly Insider BriefTighter analysis you can share internally
  • Extra audio segmentsCut interviews, behind-the-scenes commentary, quick competitive notes
  • Become an Insider for $12/month

    Leadership Access

    $149/month

    Perfect for MSPs and Vendors that run a team and need the extended tactics, executive summaries, and weekly alignment brief

  • All Insider Access benefits plus . . .
  • Invite your teamIncludes access for 5 team members with option to add more
  • Vendor Strategy BriefsThe entire library, plus new analysis every month
  • Channel ChatterAll topics, full insights, complete vendor discussion + sentiment list
  • Quarterly State of the Channel Briefing
  • Monthly AMA submission priorityAsk Dave direct questions, and skip the line
  • Get the Leadership Edge for $149/month

    Vendor Partner

    $500/month

    Perfect for channel companies or vendors looking to deepen their engagement with the show.

  • All Leadership Access benefits plus . . .
  • Get highlighted as a show sponsor You'll get placement in the show notes, throughout the website, and on our dedicated sponsors page.
  • Enjoy regular shout outs You'll be featured in a rotating format during the show
  • Become a show sponsor for $500/month

    Search all stories