Kaseya dropped their agentic AI platform at Connect 2026. Built on a billion-plus help desk tickets, 3 exabytes of backup data, and 17 million managed endpoints. The headline promise: 5x capacity without hiring a single person.
If you're an MSP owner watching your labor costs eat 50-60% of revenue while IT unemployment sits at 2.8%, that sounds like the answer to a prayer.
It's not. It's an exposure event.
5x capacity to do what?
Here's the question nobody on stage asked: 5x capacity to do what, exactly?
If you're doing 5x more password resets, 5x more printer troubleshooting, 5x more "have you tried turning it off and on again" -- you're 5x more replaceable. You're running faster on a track that leads to a cliff.
The Kaseya Digital Specialist for Ticket Triage eliminates up to 80% of categorization errors and routes tickets autonomously. That's real. MSPs adopting AI-driven PSA tools are already reporting 40-60% drops in ticket volume. End-to-end remediation handles 50-70% of L1/L2 tickets without a human touching them.
That's not a future prediction. That's happening right now across the channel.
So what happens when the commodity work disappears? Your clients look up from their invoices and ask a very uncomfortable question: what am I actually paying you for?
Kaseya already answered that question internally
While Kaseya was building the platform that promises MSPs they'll never need to hire again, they cut 450 of their own employees across four rounds of layoffs since April 2024. Sales, marketing, admin -- the roles that automation absorbs first.
They're not being hypocritical. They're being consistent. They built a system that eliminates commodity work, and then they eliminated the people who were doing commodity work.
That's the thesis. They're living it. The question is whether you're paying attention to the demonstration or just listening to the sales pitch.
The trust math doesn't help
Here's where it gets uncomfortable. Reddit MSP communities score Kaseya at -41.7 sentiment -- the lowest of any major vendor in the channel. Over half of all mentions are negative. Only 11% are positive. For comparison, their closest competitors score +22 to +26.
High watermark billing that only ended in December 2025. Multi-year auto-renewal lock-ins. Product stagnation after the Datto acquisition. Four rounds of layoffs while refinancing $4 billion in private equity debt.
MSPs are being asked to hand their entire operational model to their least-trusted vendor. That's not a technology decision. That's a risk decision. And the risk calculus matters.
The race to the bottom accelerates
169 publicly announced MSP M&A deals happened in 2025. Private equity showed up in 69% of them. The channel is consolidating into a handful of mega-platforms, and each one is racing to automate the same commodity services.
When everyone has the same agentic AI platform triaging the same tickets, what differentiates you? Price. That's it. And you can't win a price war against a PE-backed platform with $1.5 billion in annual recurring revenue.
52% of MSPs already report hiring and talent shortages as their top internal concern. One in four can't find enough cybersecurity talent. The labor squeeze is real. But the answer isn't automating the commodity work and calling it a day. The answer is redirecting the capacity you free up toward work that can't be commoditized.
Integration is the moat
Inference costs for the large language models powering these agentic systems dropped over 90% in two years. The agent execution layer -- the part that triages tickets and contains threats -- is commoditizing fast. Open-source agent frameworks are everywhere. Every platform vendor will have roughly equivalent capabilities within 18 months.
But integration? Wiring agentic systems into a company's specific processes, their exceptions, their operational history, their compliance requirements? That doesn't ride the cost curve down. It requires someone who understands both the technology and the business.
That's the layer that agentic AI platforms can't replicate. Custom builds. Cross-platform workflows. Vertical-specific automation. The stuff that requires a technical partner, not a platform upgrade.
Gartner predicts over 40% of agentic AI projects will be canceled by the end of 2027. Not because the technology doesn't work -- because rushing to deploy agents without doing the integration work produces expensive failures.
The MSPs that win this
The freed capacity from agentic AI is valuable. But only if you point it somewhere defensible.
White-label technical services become MSPs' second or third largest revenue category within 18 months of launch, at 30-50% margins on monthly recurring revenue. 55% of MSPs project double-digit revenue growth in 2026. SMBs are expected to channel $90 billion or more in new managed IT spending through 2026.
The money is there. The question is whether you're positioned to capture it as a technical partner or compete for it as a commodity vendor.
The MSPs that come out of this transition stronger will be the ones who took the capacity that agentic AI freed up and redirected it toward custom integrations, strategic builds, and partnerships that require actual engineering -- not just platform administration.
The ones that don't will discover that 5x capacity to do commodity work just means they're 5x more exposed when the next vendor offers the same thing for less.
Where we fit
We work with agencies and MSPs as an invisible technical partner. Your brand, our engineering. Custom builds, integrations, and the kind of work that agentic AI creates demand for but can't deliver on its own.
We've been building AI automation systems, custom tooling, and production infrastructure for years. Our own operations run on a proprietary system called LTFI that lets two people deliver what would normally require a team of 10-14. We didn't just adopt AI tools -- we built them.
If you're an MSP trying to figure out what to do with the capacity that agentic AI is about to free up, that's a conversation worth having. First one is free. No commitment. kief.studio/contact