A recent report from the tech talent firm Draup indicates that by 2027, more than 40% of IT skills may become partially obsolete due to the rapid adoption of artificial intelligence within IT departments. Vishnu Shankar, Vice President of Data and Platform at Draup, noted that while tech jobs are not disappearing, the nature of work is changing significantly, with AI compressing skill cycles and reshaping roles faster than many organizations anticipated. As companies modernize their toolsets, demand for AI-specific roles has surged, while overall demand for broader categories of tech talent has remained flat. Despite potential job losses—Forrester predicts AI could eliminate about 10.4 million jobs by the end of the decade—the net effect of AI adoption is expected to create 78 million new jobs. J.P. Gownder, Vice President and Principal Analyst at Forrester, emphasized the importance of investing in people who will leverage AI to enhance productivity and employee experience.
Why do we care?
If you hear “40% of skills becoming obsolete” and your reaction is to accelerate automation to protect margins, you are walking into a governance trap. And the moment something goes wrong, the client doesn’t ask the vendor, and they don’t ask the model. They ask you.
MSPs are absorbing responsibility for systems they don’t control. AI ticket triage, automated remediation, security analysis—all of it works until the edge case hits. And when it does, the technician on call is suddenly supervising a system they didn’t design, trained on data they can’t audit and making decisions they can’t fully trace.
That’s not a skills gap. That’s an authority gap. And there’s a timing problem layered on top of that. Most MSP contracts, pricing models, and SLAs were written for a pre-AI operating reality, while the tools underneath them are changing quarterly. That lag—often 12 to 36 months—means MSPs are assuming new forms of responsibility long before they’re contractually allowed to charge for them or limit their exposure.
The real risk isn’t job loss—it’s margin erosion paired with liability expansion. MSPs who don’t reprice and restructure around AI governance will end up delivering “premium outcomes” on commodity contracts. Clients will assume AI makes things cheaper; vendors will position AI as hands-off; and MSPs will be left holding the operational risk.
This matters now because AI is being adopted tactically, not strategically. The longer MSPs delay redefining roles, contracts, and client expectations, the harder it becomes to claw back control. AI can absolutely increase productivity—but only if someone is explicitly responsible for when it doesn’t behave as expected.
If MSPs get this wrong, the harm isn’t abstract. It shows up as client escalations, burned-out staff, and services that look profitable on paper until the first serious failure.