AI has made build versus buy interesting again.
A small team can now produce internal tools that would have been hard to justify a few years ago. A lightweight CRM. A reporting helper. A workflow that replaces a subscription nobody loved anyway. The cost of the first version has fallen, which means some SaaS products will have to defend value they used to inherit by default.
That does not mean SaaS is dead.
In an enterprise, buying software is often about transferring risk. Compliance, uptime, support, vendor liability, security review, procurement, and maintenance all matter. The custom tool that took three days to generate can become a permanent obligation the moment it touches regulated data or core operations.
This is where shadow AI gets messy. Employees will build or stitch together tools because the official path is too slow. Sometimes that is creativity. Sometimes it is unmanaged risk with a better demo.
The line has shifted, though. Leaders should stop treating build versus buy as a stale procurement ritual. The right answer depends on scale, ownership, data sensitivity, and how much future maintenance the organization is honestly willing to carry.
Generating code is cheaper now. Owning software is not.
The gap between those two facts is where the decision lives.
Related episode: Is SaaS Dead? Build vs. Buy in the Age of AI.
