A Microsoft VP is pushing back on the idea that SaaS is dying. Instead, the argument goes, SaaS is turning into the operating system for AI agents — and that shift changes how companies should think about their software stack.
The claim matters because it lands at a moment when many companies are stuck. They are moving from AI pilots to real deployments and seeing measurable results. Teams are generating insights, completing research, and drafting work in a fraction of the time. What took months can now take days. What took days can now take minutes.
Why AI Momentum Slows Down in Companies
Then the momentum slows. That is the moment companies learn that productivity gains alone do not automatically translate into business transformation.
The numbers back this up. McKinsey finds that 62% of organizations are experimenting with AI agents, but only 39% report an earnings impact. Gartner expects more than 40% of agentic AI projects to be canceled by the end of 2027.
The reason is simple: one critical part of the workflow hasn't changed. Organizations are still relying on people to bridge the gaps between tools and tasks.
What "Operating System for AI Agents" Actually Means
The Microsoft VP's framing suggests the next phase of SaaS is not about replacing software with AI. It is about software becoming the layer that coordinates AI agents — the place where agents get their instructions, their data, and their guardrails.
In that view, SaaS does not disappear. It becomes the foundation agents run on. That is a different bet than the "SaaS is dead" narrative that has circulated as AI models get more capable.
Our Take: The Real Problem Is the Workflow, Not the Software
In our view, the Microsoft VP is right about the direction but the harder truth is buried in the McKinsey and Gartner numbers. The gap between 62% experimenting and 39% seeing earnings impact is not a software problem. It is a workflow problem.
To put it plainly: if companies keep using people to bridge the gaps between AI tools, no amount of agentic capability will fix the earnings gap. SaaS becoming the operating system for AI agents only helps if organizations actually redesign the work around it.
For readers, the practical takeaway is this: do not buy the "SaaS is dying" story, and do not buy the "AI will fix everything" story either. The companies that win will be the ones that treat SaaS as the coordination layer for agents — and rebuild their workflows so agents and people are not stepping on each other.
Until that happens, expect more pilots, more canceled projects, and more headlines about AI momentum slowing down.
