The SaaS Reset: What the AI Panic Reveals About Software’s Future

Episode 13

February 06, 2026

Over the past week, one narrative has dominated tech and market coverage.
In just a few trading sessions, software and data companies shed hundreds of billions of dollars in market value as investors reacted to a new wave of agentic AI releases.

AI is no longer just enhancing software — it’s starting to question whether some software needs to exist at all.

A wave of headlines followed the release of new agentic AI capabilities from companies like Anthropic, sparking a sharp sell-off across software and data companies. Business media quickly framed the moment as a potential “SaaSpocalypse”— a dramatic label, but one that reflects a real and growing tension in the market.

Recent coverage across business and financial media suggests investors are reassessing a long-held assumption: that subscription-based software is inherently defensible and resistant to disruption.

This week suggests otherwise.

What happened (and why it matters)

The immediate trigger was simple: AI tools are moving fast — from copilots to autonomous agents capable of drafting contracts, conducting legal research, generating reports, and orchestrating workflows that previously justified entire SaaS categories.

The market reaction was swift:

  • Legal, data, and information service companies were hit particularly hard

  • Broader software indices followed

  • Concerns spilled into the wider tech sector

In some cases, single-day declines reached double digits across sectors long considered among the safest in enterprise software.

But this didn’t start this week.

For months, AI companies have been rolling out tools that allow non-technical users to assemble software, workflows, and internal systems with minimal engineering effort. Whether these tools are perfect is almost irrelevant.

What matters is that they are good enough to challenge long-standing assumptions about software pricing, vendor lock-in, and the need for large, rigid platforms in the first place.

The capital side has noticed.
As Apollo Global Management and other private credit firms have quietly signaled by reducing exposure to software-heavy portfolios, the money is already adapting to this new reality.

A needed reality check

Despite the panic, this isn’t the end of software.

Industry leaders, including Jensen Huang, have repeatedly pointed out that AI is unlikely to replace complex enterprise systems outright. These systems are regulated, deeply embedded, and tied to real-world operational constraints.

But something is ending.

What’s under pressure is software that:

  • exists mainly as a thin interface

  • charges high recurring fees for generic functionality

  • is difficult to adapt, integrate, or evolve

In short, software built to monetize friction rather than eliminate it is facing a structural problem.

In an AI-first world, costly and rigid vendor software is becoming harder to justify.

Our take: this isn’t a crash — it’s a reset

From our perspective, this moment isn’t about panic or hype.
It’s about repricing assumptions.

AI is compressing the value of generic software and shifting advantage toward teams that can build faster, integrate AI directly into core workflows and tailor systems to how their business actually operates

In this environment, the question for many teams is no longer which vendor to choose — but whether buying off-the-shelf software still makes strategic sense at all.

This is exactly why we built our AI Velocity Pod.

Instead of selling another off-the-shelf tool, we work with teams to:

  • design and ship custom AI-powered systems, quickly

  • replace fragmented SaaS stacks with lean, purpose-built software

  • reduce long-term dependency on expensive vendors and licenses

  • move AI from experimentation into production-grade reality

The goal isn’t speed for its own sake.
It’s building software that’s cheaper to run, easier to evolve, and harder to commoditize.

The real signal from this week

AI isn’t killing software.
It’s forcing a long-overdue question:

Should this be bought — or should it be built?

Teams that learn to build well, with AI embedded from day one, won’t just weather this shift.
They’ll outgrow competitors still locked into slow, costly vendor ecosystems.

That’s the opportunity hiding beneath this week’s headlines.

Sources & Further reading

Exploring ways to move faster while reducing dependency on costly vendor software?
Let’s start a conversation.

P.S. We want to make sure this newsletter hits the mark. So reply to this email and let us know what you think.