There has been a growing narrative in tech that AI could wipe out traditional software.
Some are even calling it the “SaaSpocalypse.”
But the CEO of AWS is not buying that story.
Matt Garman is taking a very different stance on where things are heading.
Instead of seeing AI as a threat to software as a service (SaaS), he sees it as an expansion layer, one that could make the market even bigger.
In his words, this shift represents a “huge business opportunity” for AWS, especially as AI agents begin to take on more real work inside companies.
That perspective explains AWS’s latest move.
The company is no longer just focused on cloud infrastructure. It is moving up the stack into actual software products, building tools designed for everyday workers, not just developers or engineers.
One example is a new desktop app that allows users to interact with an AI assistant to handle tasks like creating presentations or scheduling meetings. Alongside that, AWS is rolling out specialized AI applications for hiring, healthcare, and supply chain operations.
That is a big shift.
For years, AWS has been the invisible layer powering other companies’ software. Now, it is starting to build the software itself.
And AI is the reason why.
Because once AI agents can perform tasks instead of just supporting them, the line between infrastructure and application starts to blur.
That is where the SaaS debate comes in.
Some investors believe AI agents will replace traditional software interfaces entirely. Instead of using tools like CRMs or dashboards, users could simply tell an AI what they want, and it handles everything in the background.
But Garman is pushing back on that idea.
He is betting that companies will still buy software, just in a different form. Instead of static tools, they will be dynamic systems powered by AI, where agents handle workflows, automate decisions, and integrate across platforms.
In other words, SaaS does not disappear.
It evolves.
There is also a strategic layer behind this.
AWS recently expanded its partnership with OpenAI, allowing businesses to access OpenAI models directly through its cloud platform, something that was previously limited due to Microsoft’s earlier exclusivity.
That move signals a broader ambition.
AWS does not just want to host AI.
It wants to be the place where companies build, run, and deploy AI powered businesses.
And that includes competing more directly with traditional SaaS providers.
But the shift is not without tension.
If AI agents become capable enough, they could reduce the need for complex software interfaces. That would challenge how SaaS companies price, package, and deliver their products.
At the same time, building and maintaining AI powered systems is expensive, which could push companies toward platforms like AWS that already have the infrastructure in place.
So both things can be true at once.
AI can disrupt parts of SaaS while also expanding the overall market.
That is the balance Garman is pointing to.
What this moment really shows is how the conversation is changing.
It is no longer about whether AI will replace software.
It is about what software becomes when AI is built into its core.
And that is a much bigger shift.
So the real question is not whether SaaS will survive.
It is whether the companies building it today can adapt fast enough before platforms like AWS start owning more of the stack.
