Workday lays off 500 people amid the "SaaSpocalypse": it denies AI played a role, the market doesn't buy it

By Zendoric · September 30, 2026. Workday is laying off around 500 employees — 2.5% of its workforce — mainly in its product and technology team, the company disclosed on Tuesday in a filing with the U.S. Securities and Exchange Commission (SEC). It is the second round of job cuts this year for the HR and finance software maker, which had more than 21,000 employees at the end of January. Workday also cut its operating margin guidance for the fiscal third quarter and for full-year 2027, because of restructuring charges stemming from the layoffs.

The company had already laid off some 400 people in February. Barely a week later, its then chief executive, Carl Eschenbach, stepped down; he was succeeded by Aneel Bhusri, Workday co-founder and executive chairman, who thus took the reins of the company he founded for the fourth time. Workday shares are down roughly 12% so far this year.

The backdrop is what the industry has dubbed the "SaaSpocalypse": the mass sell-off throughout 2026 of shares in subscription enterprise software (SaaS) vendors, driven by investor fears that AI will let companies build their own in-house tools — or buy them from cheaper startups — instead of paying licence fees to giants such as Workday. That fear, according to Business Insider, has wiped hundreds of billions of dollars of market value from the sector so far this year.

Workday, however, has not cited AI as the reason for either round of cuts: the company attributes the move to the need to "better align team structures with Workday's strategic growth priorities," and says it will keep hiring in strategic areas and locations during fiscal year 2027. Bhusri, moreover, has been defiant in the face of the replacement narrative: on the August earnings call he said he had not spoken to "a single customer" looking to replace Workday with something built in-house or bought from a startup, and that a quarter later the situation was unchanged.

It is worth separating fact from narrative here. What is verifiable is this: a company with more than 21,000 employees is cutting 2.5% of its workforce in its second round of layoffs in eight months, is trimming margins and has lost its CEO mid-process. What is not verified — and Workday does not confirm it — is that AI is the direct cause. Blaming every tech layoff of 2026 on AI would be simplistic; but ignoring that the company's share price, and those of its peers, is indeed being repriced by the market on the basis of a narrative about AI — whether or not it is already true — would be just as naive.

That distinction between what AI does today and what investors fear it will do tomorrow is the key to the "SaaSpocalypse." No company needs to have already replaced its Workday systems with AI-generated code for the market to punish subscription software vendors: it is enough that there is a credible possibility of it happening in two or three years. The layoffs, in that sense, read less as a response to a substitution that has already taken place than as pre-emptive cost discipline, getting ahead of a risk premium that capital markets are already charging today.

This fits with what we have been observing in our coverage of tech employment sector by sector: in technology, the workforce does not win or lose as a block; those who bring architecture, AI governance and integration win, and the most redundant roles within the product teams themselves lose. That Workday's cuts are concentrated precisely there — and not, say, in sales — suggests a reorganization toward flatter structures underpinned by AI-assisted development tools, even if the company does not frame it in those terms.

In the short term, this is exactly the kind of adjustment that should not be played down: these are people with real paychecks and real careers, in a second round in less than a year, under leadership that has wobbled. But in the medium term, if competitive pressure from AI forces incumbent software vendors to become more efficient — and, above all, if it genuinely makes it cheaper to build custom software for companies that today cannot afford licences like Workday's — the net result is more tools available at lower cost to more organizations. That is the paradox of this phase of AI applied to the enterprise: it hits those who sell the software first, before potentially making it cheaper for everyone who uses it.