In a recent analysis on the Schnitger Corporation blog, author Monica Schnitger argues that software-as-a-service (SaaS) isn’t about to vanish simply because artificial intelligence is advancing rapidly. Commentary in the tech industry often predicts AI will replace traditional SaaS tools with custom, AI-generated applications tailored to a company’s needs. But that outlook overlooks significant realities about software delivery and economics.
One cornerstone of Schnitger’s argument is the scale and complexity involved in business software. Commercial SaaS vendors build systems to meet the needs of hundreds or thousands of users, with teams of professional developers, support staff, security specialists, trainers, and partners developed over the years. These vendors amortize development and maintenance costs across large customer bases, keeping per-seat prices manageable and delivering reliability that bespoke AI solutions can’t match yet.
Schnitger also notes that using AI to generate custom tools isn’t free or trivial. AI training and execution require expensive infrastructure, and API calls to generative models remain costly at scale. She cites examples where building even simple apps with AI can cost far more than expected, especially as requirements grow or change, making in-house AI-driven software economically unattractive for many organizations.
Discussion in broader tech coverage supports the idea that AI reshapes rather than replaces software models. Some enterprises are indeed reallocating budgets from traditional apps to AI tooling, and investors worry about the “SaaSpocalypse.” Still, industry analysts and providers are adapting by embedding AI capabilities into existing SaaS offerings rather than abandoning the model.
The bottom line from Schnitger’s perspective is that SaaS has deep practical and economic roots in enterprise IT. AI will change how software is built and consumed, but it won’t make SaaS obsolete. Vendors that integrate AI thoughtfully into their platforms and continue to deliver scalable, supported services are likely to remain relevant even as custom, AI-generated tools emerge alongside them.