Rich Freeman in Channelholic with the evolving landscape of managed services, focusing on the distinction between “systems of record” and “systems of action.” Systems of record are core business applications like ERP and CRM software that store critical data, while systems of action automate tasks entirely, using advanced AI capabilities. According to Omdia, sales for agentic AI software are projected to increase significantly, from $1.5 billion last year to $41.8 billion by 2030, showcasing a 175% growth rate over five years. In contrast, traditional generative AI is expected to grow at a comparatively lower rate of 90% over the same period. The article highlights that incumbent vendors must adapt to this shift or risk losing market relevance as newer, more agile competitors emerge.
Over in the New Stack, A recent survey highlighted by Kevin Reeuwijk, a distinguished architect at Spectro Cloud, underscores the need for human oversight in AI operations, as even a 90% accuracy rate in AI actions could lead to significant errors. This shift towards operational AI frameworks is evident, with a 25% rise in AI adoption correlating with a 7.5% improvement in documentation quality, according to Google’s annual DevOps Research and Assessment report. The article emphasizes the importance of creating a unified AI and automation layer to streamline incident management, as well as the emergence of ModelOps to better integrate AI into software development processes. Furthermore, the potential of multimodal AI, which combines various forms of input for enhanced diagnostics, could significantly improve operational efficiency.
And also from the New Stack, “Why Your Dashboards are Obselete in the Age of AI”. Companies like Amazon and McDonald’s are investing heavily in automation, with projections indicating that global AI spending could exceed $2 trillion by the end of 2026. Yet, as autonomous agents and robots generate vast amounts of data, reliance on outdated dashboard technology can lead to costly delays and mismanagement. Experts argue that organizations must transition to a new model of machine-to-machine business intelligence, where systems communicate and make decisions autonomously, leaving human oversight to strategic governance rather than operational micromanagement. Companies that adapt to this shift early by consolidating data and embedding intelligence into workflows will likely lead the way in operational excellence, setting a new standard for the future of enterprise management.
Why do we care?
Everyone likes the idea of systems of action because they sound efficient. Fewer clicks. Fewer people. Faster outcomes. But systems of action don’t just do work—they make decisions that used to belong to humans.
That’s why agentic AI is growing faster than generative AI. Not because it’s safer or smarter, but because it promises leverage. And leverage always shifts risk.
If you’re an MSP and you help a customer move from dashboards to autonomous execution, you’ve just crossed a line—from reporting on systems to authorizing behavior. Most MSPs haven’t updated their contracts, pricing, or governance models to reflect that reality.
The New Stack is right that dashboards are losing operational relevance. But they’re wrong if anyone thinks that means less responsibility. It means more—just earlier and quieter.
And here’s the key connection: systems of record were forgiving. Systems of action are not. You could misunderstand a dashboard for months. You can’t misunderstand an autonomous system without consequences.
This matters now because vendors are racing toward action-first platforms. MSPs who follow without redefining authority will discover too late that they didn’t just deploy software—they accepted accountability.
The real opportunity isn’t agentic AI.
It’s being the provider who knows exactly where automation should stop—and can explain why.
CES didn’t show us a future of smarter tools—it showed us a future where someone has to be accountable when automation acts, and MSPs are closest to the blast radius.