AI isn’t replacing SaaS—it’s transforming how businesses interact with it. Instead of navigating multiple applications, employees will increasingly rely on AI to translate intent into action, using SaaS systems as the underlying infrastructure. This shift moves SaaS into the background as “plumbing,” while AI becomes the intuitive interface or “faucet” that delivers outcomes on demand.
For organizations, the real challenge isn’t adopting AI tools—it’s fixing fragmented SaaS environments, improving data integration, and building systems that can work together seamlessly. Without strong foundations, AI initiatives fail to deliver meaningful results.
CIOs must now take on a more strategic role, focusing on three priorities: strengthening core systems, building AI as an orchestration layer across the enterprise, and redesigning how people work. Companies that embrace this shift toward intent-driven operations will gain a lasting competitive advantage, while those that layer AI onto outdated systems will fall behind.
“Software is eating the world.” That simple phrase turned out to be an incredibly visionary observation, made at exactly the right moment. At the time, an investor or business leader using that as a central tenet of their approach to business building and digital transformation in the following years likely would have performed pretty well.
So it’s inevitable that almost two decades on, the “AI is eating software” narrative is gaining momentum. It’s neat, it’s memorable, and the two waves of innovation share many characteristics. But I believe this simple framing deserves scrutiny before it hardens into conventional wisdom.
We have to ask ourselves some hard questions and really think about the answers. Is AI really killing SaaS? Is the era of cloud applications ending? And are the platforms enterprises have spent the last decade building their operations around really about to be completely displaced?
These questions matter because big investment decisions for large companies depend on the answers. Let’s look at what the evidence shows so far.
Is AI Really “Eating SaaS”?
Bottom line up front: SaaS isn’t disappearing. In many respects, it’s becoming more essential than ever. AI isn’t changing the existence of SaaS platforms, but it’s likely to change how we interact with them.
That means that the real shift underway isn’t from SaaS to something else, but from operating systems to expressing intent, and from navigating screens to achieving outcomes. Understanding that distinction is the starting point for any serious conversation about AI leadership and enterprise transformation.
To see where we’re headed, we need to be honest about where we are now. SaaS has truly changed how businesses operate, but it has also brought a downside: too many tools. Most large companies now use over a hundred different SaaS apps. Each solves a problem, but together they create a lot of fragmentation. Employees spend too much time switching between systems and managing tools, rather than focusing on their actual work.
The cognitive load this places on knowledge workers is one of the most significant drags on organizational productivity. The total weight of SaaS tools doesn’t optimize for experience or flow, leaving humans as operators of systems. And operators spend their time managing the machinery, not driving the outcomes.
The Big Shift: From Plumbing to Faucets
It may not be a perfect comparison, but thinking about SaaS and AI like plumbing can help explain things.
- SaaS is the plumbing. It’s largely hidden “behind the wall” in the pipes, valves, pressure systems, and the infrastructure that moves data and transactions (or water) through the enterprise. It’s essential, complex, and largely invisible when it works well. No one thinks about the plumbing when they turn on a faucet. They think about the water.
- AI is the faucet. It’s the simple, intuitive interface to an extraordinarily complex underlying system. People don’t want to manage pipes. They want outcomes on demand, and they want them without needing to understand the infrastructure that delivers them.
So, for enterprises, AI becomes the interaction layer, or put another way, the “faucet” through which employees access the full capability of the enterprise’s SaaS infrastructure. The SaaS platforms remain the execution layer, providing the plumbing that stores data, enforces compliance requirements, and processes transactions.
AI as the New Interface: From Navigation to Intent
The old way of using business software is familiar and, after years of use, ready for an update. Employees log in, find the right module, search for records, click through forms, export data, and finally get their work done. In this setup, people do most of the thinking.
The intent-driven model flips this around. Employees state what they want to achieve, and AI handles the behind-the-scenes work. For example, a sales leader who needs a proposal with risk modeling doesn’t have to jump between different systems. They just say what they need, and AI leadership puts it together.
It’s important to see that this isn’t just a better chat interface; it’s a full update to how work gets done. For real business transformation, this difference matters. Companies that just add a chat layer to old workflows will get some convenience, but those that redesign their processes around intent-driven AI will gain a lasting edge.
SaaS Is Becoming Critical Infrastructure
If the framing above is correct, then the prediction that AI will kill SaaS, largely fed by the dramatic stock price movements of publicly listed SaaS companies, misunderstands the dynamic entirely. SaaS doesn’t become less important as AI matures; it becomes more important, but in a different way.
Likely, the core systems SaaS platforms provide aren’t going anywhere. They hold the data AI leadership needs, enforce compliance rules, and handle the transactions that keep businesses running. Replacing all this with AI-only systems isn’t practical or needed. The plumbing already works. What needs to change is the interface.
In practice, this means reframing the enterprise technology stack. SaaS becomes mostly hidden infrastructure that’s essential, managed for reliability and integration quality, but increasingly invisible “behind the wall” to the end user. AI becomes the orchestration and experience layer, or the primary surface through which employees interact with the enterprise’s capabilities. And competitive advantage shifts correspondingly. The question becomes how seamlessly those tools are orchestrated and how quickly and fluidly the intelligence layer can translate intent into coordinated action across the underlying systems.
Why Most AI + SaaS Strategies Fail
So why are so many AI projects struggling? While there isn’t a lot of hard data, I believe the main problem is that companies add AI on top of a messy SaaS setup and expect big changes. Instead, they get scattered AI results, incomplete insights, broken automation, and advice that employees soon stop trusting.
This suggests the real issue is weak plumbing. AI is only as good as the data it can access, and the data available to most enterprise AI systems is limited by the quality and integration of the underlying SaaS infrastructure. Lack of integration between platforms means AI agents can see part of the picture but not the whole, and can take action within one system but can’t orchestrate across workflows spanning several systems.
There’s a clear (and perhaps a little blunt) underlying principle here: if the plumbing is broken, the faucet won’t work. For AI leadership and those responsible for capital allocation decisions, that means no amount of investment in the AI layer will compensate for a fragmented, poorly integrated SaaS foundation underneath it.
The Rise of the Intent-Driven Enterprise
The organizations getting this right share a set of structural characteristics worth examining. Work in these enterprises starts with intent, not systems. Employees are oriented around outcomes and express what they need, and the AI layer orchestrates across the underlying platforms required to deliver it. AI agents coordinate across systems in real time, executing actions without waiting for manual handoffs at each step.
This is underpinned by an integrated data environment where the AI layer has consistent, governed access to the information it needs across the SaaS ecosystem, enabling real-time decision-making.
This shift also changes organizations in a big way. Employees stop being system operators and become outcome designers. They define their goals and work with AI leadership to determine the best way to achieve results. Companies that redesign how people and AI work together are building the skills needed for long-term success.
The CIO Mandate Is Being Rewritten
The CI’s role has always changed with technology. But today’s shift is more strategic and more challenging than before, and that brings a real opportunity.
The new mandate is to be the architect of enterprise experience and intelligence. That requires leading in three distinct directions simultaneously.
1. First, strengthen the foundation. This means cutting out extra SaaS tools, improving data quality, and setting standards for systems to work together. It’s not glamorous, but it’s essential plumbing work that can’t be put off.
2. Next, build AI as the orchestration layer. Design workflows that let AI agents work across all systems, not just within one platform, so they can handle tasks from start to finish.
3. Finally, redesign the human experience of work. The CIO who leads this change with clear communication, empathy for employees, and a real focus on helping people grow, not just cutting costs, will drive lasting transformation.
What This Means for Competitive Advantage
I don’t think the companies that will lead the next era will be the ones with the fanciest models or the biggest AI budgets. Instead, it will be those who can consistently turn intent into action across their entire business.
Two paths are opening up. The first is AI as enhancement: layering intelligence onto existing workflows and tools to generate incremental productivity gains. This is the path of least resistance, and the one most enterprises are currently on. The returns are real but modest in the long term, and they’re available to every competitor with access to the same tools.
The second path is AI plus operating model redesign: restructuring how work is designed, how systems are integrated, and how human capability is deployed around an intent-driven architecture. This path is harder and slower in the short term, but it generates a structural edge. That kind of advantage compounds over time and is genuinely difficult to replicate.
What Leaders Should Do Now
AI leadership that wants to move from the first path to the second has some clear steps to follow. Start by simplifying your SaaS setup. Remove extra tools, improve integration, and invest in clean, unified data. This is the must-have foundation for reliable AI.
Next, focus on the orchestration layer. Design workflows so AI can coordinate across systems, instead of just adding AI to existing silos. Think in terms of flows and outcomes, and use this approach in selected areas to show real progress and value to your team and leaders.
The last point is the most human. Most CIOs know employees don’t care about system architecture or which vendor’s API is used. What matters to them is how easily they can get their work done and how quickly they can turn their goals into results.
That experience will become one of the most significant talent and productivity differentiators of the coming decade. The future of enterprise value will be defined by how effortlessly human intent turns into execution, and the most effective AI leadership will understand and act on that.
Key Takeaways
Companies and AI leadership that commit to fixing the basics, building a strong orchestration layer, and truly redesigning how people work will reach a whole new level of efficiency and intelligence.
The gap between those enterprises and the ones still navigating their way through a hundred disconnected tools will likely widen. But the window to be on the right side of that gap is wide open. The architecture decisions being made right now about data, integration, orchestration, and experience are the ones that will define competitive position for the better part of the next decade.
The faucet is ready. The real question is whether the plumbing underneath can handle it.
