Paul Appleby is Chief Executive Officer of Virtana, one of the World’s Deepest and Broadest Observability Platforms.
One of the great ironies of our time is that AI, expected to replace not only enterprise software companies but even work itself, will be the first to tell you that reports of SaaS’s death are greatly exaggerated. The broader industry consensus, echoed by various tech executives and analysts, is, “SaaS is not dead, but it is undergoing a massive, AI-driven evolution.”
I am not an AI cynic. On the contrary, I think AI could be more significant than the Industrial Revolution or the shift from a manufacturing economy to the service economy. As the CEO of a SaaS company, I simply disagree that the dawn of AI spells the death of SaaS.
We have been here before. The book was not replaced by the radio, the radio was not replaced by the television, the television was not replaced by the computer nor was the computer replaced by the smartphone. Instead, each new invention led to adaptation and innovation.
AI will not replace SaaS; it will improve it. Replacement is binary. Instead, we should think of AI as a reconstitution of how we engage with our software.
The old SaaS model is under pressure.
The problem is that businesses across the tech industry are still doggedly asking a yes-or-no question: Is SaaS dead? The resulting market volatility has created a kind of Schrödinger’s SaaS. SaaS is, like the cat in Schrödinger’s thought experiment, simultaneously dead and alive.
In 2025, investors began reassessing software companies as AI-native tools demonstrated that some traditional software functions could be delivered faster, cheaper or with far less human intervention. Software stocks were hit particularly hard during this time, fueled in part by concerns that AI agents could disrupt the per-seat pricing model that has underpinned much of SaaS.
Then, in early 2026, nearly $1 trillion in enterprise software valuation (registration required) was wiped out, according to Reuters, with the selloff driven by growing concerns that AI could fundamentally disrupt traditional SaaS business models. The episode became known as the “SaaSpocalypse.”
But the story is more nuanced than a wholesale rejection of software.
As we are nearing the end of 2026, the picture looks considerably more nuanced. Stripe reports that SaaS equity prices have returned to roughly their pre-SaaSpocalypse levels, while transaction data from the 100 largest non-AI SaaS companies on its platform showed only a brief dip followed by a swift recovery and continued growth. At the same time, new SaaS platform businesses on Stripe are up more than 180% year over year.
What is changing is the economics and expectations around SaaS. AI is challenging the economics and expectations that have defined SaaS for years. AI agents can perform tasks that once required people to interact directly with software. AI coding tools are also making it increasingly possible for companies to build or customize software themselves.
Those shifts are forcing investors and software companies alike to reconsider where value will accrue in an AI-native market.
That shift is also giving rise to a new category: AI-native services. Instead of selling software that helps a person perform a task, these companies are increasingly selling the task itself as an outcome. The customer pays for the result rather than the software, and AI performs much of the work behind the scenes.
This does not make SaaS obsolete. It changes where the software sits in the value chain. Software can become the infrastructure, intelligence layer or system of record powering an outcome that the customer ultimately cares about. In some cases, the software remains visible. In others, it effectively disappears behind an AI agent or service.
As with all technological change, software as a service will continue to exist, but the model will shift to adapt to, in this case, AI. In the process, some SaaS products will die off, but many will emerge as more powerful tools.
As with buggy whips and automobiles, Blockbuster and Netflix, taxis and Uber, if existing players do not evolve to incorporate new technology where relevant, the new technology will displace them. Whether or not they survive depends on solving the problems in the industries they serve.
SaaS is the foundation for AI-driven outcomes.
Let me pose the question of “Is SaaS dead” another way: What business problem does AI solve? No one’s business problem is “SaaS.”
A healthcare organization’s problem could be that their connected medical devices cannot fail while in use. A government agency’s problem might be that their emergency-response platform must remain stable and responsive in a crisis. A business may want to understand where the checkout process lags and customers drop off.
AI can help solve these problems. To do so, AI does not need to build a tool to do what SaaS does. There is already a trusted, integrated SaaS platform handling it.
It would, however, be incredibly powerful if a SaaS platform could use AI to automate the software services people now have to manage manually. The dream is to free employees from SaaS platform management so they can work on crisis response, connected device fail-safes or a quick and easy checkout experience.
The promise of AI is real-time reasoning and intervention, but premised on a trusted, reliable, enterprise-grade SaaS platform that orchestrates, governs and drives efficiency.
This is why as businesses incorporate AI, there will not suddenly be less work to be done, but rather, more business opportunities to pursue.
SaaS and AI will work hand in hand.
The next generation of enterprise software as a service will deliver trusted, integrated platforms with complete access to quality data, and capabilities AI can execute autonomously in response.
Saas and AI will work hand in glove. AI’s value will not be borne out in building software to replace SaaS, but rather, SaaS platforms that incorporate AI will be able to solve the complex problems unique to the AI era.
So, for those of you still wondering, is SaaS dead or alive, the answer is, of course, Schrödinger: both.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?