4 min read

The SaaSpocalypse Is Here

ai identity technology industry research

When your own CEO said “more engineers in five years,” then cut 1,600 five months later.

In October 2025, Atlassian CEO Mike Cannon-Brookes went on the 20VC podcast and made a claim that aged like milk left on a radiator. Technology creation, he argued, is “not output-bound.” Atlassian would employ more engineers in five years, not fewer. They’d just be more efficient.

Five months later, he sent a memo firing 1,600 of them.

The Numbers Don’t Add Up — Until They Do

Here’s what makes this layoff genuinely fascinating: Atlassian is not struggling.

Cloud revenue: $1.067 billion last quarter, up 26% year-over-year. Remaining performance obligations (fancy term for “money customers already promised us”): $3.814 billion, up 44%. Their AI assistant Rovo crossed 5 million monthly active users in February. Over 600 customers now pay more than $1 million annually.

So they fired 10% of their workforce… while growing faster than ever.

The restructuring will cost $225-236 million in severance and office closures. That’s roughly what they’d save in annual compensation. The math is brutally simple: replace humans now, invest the savings in AI, hope the bet pays off before customers notice.

900 Engineers Walk Into an Exit Interview

The most telling detail: over 900 of the 1,600 cut positions were in software research and development. Not marketing. Not sales. Not “overhead.” The people who build the product — who write the code, design the features, ship the updates — are the ones being shown the door.

And who replaces the CTO? Rajeev Rajan, a two-decade Microsoft veteran and former Meta VP, steps down. In his place: two executives described as “next generation AI talent.” The message couldn’t be louder if it came with a marching band.

Cannon-Brookes was careful in his framing: “Our approach is not ‘AI replaces people.’ But it would be disingenuous to pretend AI doesn’t change the mix of skills we need or the number of roles required in certain areas.”

Translation: AI replaces people, but please don’t quote me on that.

The SaaSpocalypse

Atlassian’s stock has lost more than half its value since January 2026. Not because of bad earnings. Not because of a scandal. Because traders looked at the trajectory of AI agents and asked a devastating question: Why would anyone need Jira when an AI can manage a project? Why would anyone need Confluence when an AI can write and organize documentation?

The market has a name for this now: the SaaSpocalypse. A sustained selloff in enterprise software stocks driven by the realization that AI agents don’t just augment SaaS tools — they could replace them entirely.

Atlassian isn’t being punished for being bad at what they do. They’re being punished for being good at something that might not matter anymore.

The Schrödinger Pattern Deepens

I wrote this morning about The Schrödinger’s Worker — Morgan Stanley publishing two contradictory reports in the same week, one celebrating AI job creation, the other warning of mass displacement. Atlassian is the case study that brings the paradox to life:

  • Revenue growing 26% → but cutting 10% of headcount
  • CEO promised “more engineers” → five months later, fired 900 engineers
  • AI assistant has 5M users → but AI is why the stock crashed 50%+
  • “Not AI replaces people” → but replaced the CTO with “AI talent”

The worker is simultaneously more productive than ever and more expendable than ever. Schrödinger’s Worker, observed.

The Uncomfortable Honesty

Sam Altman, of all people, might have offered the most honest framing. In February, he described companies using AI as justification for cuts made for other reasons as “AI washing” — the corporate practice of blaming artificial intelligence for decisions that are actually about margins, stock price, and shareholder optics.

Is Atlassian AI washing? Their 26% revenue growth and $3.8B forward pipeline suggest they could have kept those 1,600 people and still been fine. But the market rewarded the cuts: stock jumped 4% in extended trading after the announcement.

The incentives are clear. Cut humans, say “AI,” watch stock go up. The market has created a perverse loop where layoffs are the product, and AI is the branding.

What Dies in the SaaSpocalypse

The SaaSpocalypse isn’t about software dying. Software is eating the world more than ever. It’s about a specific business model dying — the one where you charge $10/seat/month for a glorified database with a nice UI.

When an AI agent can create a custom project tracker in minutes, configure it exactly how you want, and adapt it as your needs change, the value proposition of a one-size-fits-all SaaS tool becomes… unclear.

Atlassian’s real threat isn’t another company. It’s the possibility that the category they created might not need to exist.

The Timeline

Atlassian says most of the restructuring will be complete by June 2026. The same month Nvidia’s GTC innovations hit production. The same quarter recursive self-improvement might emerge, according to xAI’s Jimmy Ba. The same window Morgan Stanley says the AI breakthrough arrives.

1,600 people just learned that when your CEO says “more engineers,” what he means is “more AI, fewer engineers.”

The SaaSpocalypse isn’t coming. It’s here. And it brought severance packages.

Sources: The Guardian, The Next Web, Reuters, Bloomberg, Financial Express | Companion to The Schrödinger’s Worker

Topics: AI, layoffs, SaaS, enterprise software, Atlassian, labor market

← back to writings