The Productivity Claim That Cut 1,000 Jobs Was Never Audited

On Thursday, May 28, 2026, Wix CEO Avishai Abrahami told his company it would lose roughly 20% of its people. That is around 1,000 jobs out of a headcount that stood at 5,277 at the end of Q1 2026, per the company’s shareholder update. Reuters reported two stated reasons: a strong shekel against the dollar, and the rapid evolution of AI.

The currency pressure is real and quantified. The shekel has appreciated nearly 30% against the dollar over the past year, reaching a 33-year peak, and most of Wix’s teams are Israel-based while its revenue is largely dollar-denominated. That is a structural cost squeeze you can read off an exchange rate. It is auditable.

The AI half of the justification is not.

A Measurable Claim Nobody Measured

Read what Abrahami actually said about AI, as quoted via CNBC and Fortune. AI is “the most significant shift in how companies are built since the invention of modern programming languages in the 1970s.” The company needs “fewer layers and fewer workers” and must become “a faster, leaner, and flatter organization.”

Strip the rhetoric and a falsifiable proposition is sitting underneath: AI now produces enough of the work that 1,000 people previously produced that the company can operate at scale without them. That is not a mood. It is a measurement. Output per unit of labor either rose enough to absorb a 20% headcount cut, or it did not.

So here is the uncomfortable question for every executive watching this play. Where is the instrumentation that proves it? What is the baseline output of those 1,000 roles, what is the AI-augmented output that replaced it, and who verified the delta before the decision became irreversible?

In most organizations making AI-justified cuts in 2026, that instrumentation does not exist. The decision runs on narrative confidence, not on a measured scoreboard. A currency hedge has an audit trail. A productivity claim that ends a thousand careers frequently has none.

The Pattern Has a Name Now

Wix is not an outlier. It is the latest entry in a sequence. Block cut nearly 4,000 jobs in February 2026, close to half its workforce, with CFO Amrita Ahuja citing the chance to “move faster with smaller, highly talented teams using AI to automate more work.” Cisco cut about 4,000 people, 5% of its workforce, in May; the stock jumped 13% after the announcement. Meta eliminated around 8,000 jobs, roughly 10% of headcount, the same month.

MIT Sloan professor emeritus Paul Osterman gives the move a label. He calls it “AI washing.” Speaking to Fortune, he framed the mechanism plainly: “AI is a perfect excuse to justify big layoffs. It makes it seem as if it’s not our decision, our fault, it’s the technology.” He adds, in his account, that companies have pushed the “smaller, more productive teams” framing for 20 years. The vocabulary changed. The maneuver did not.

I want to be careful here, because the caveat is the whole point. Osterman’s “AI washing” characterization is expert opinion, not a forensic finding about Wix specifically. I am not asserting that Wix did or did not capture genuine AI productivity. The public record does not establish either. That absence is precisely the governance failure. When a claim of this consequence cannot be confirmed or refuted from anything the company has published, the problem is not the answer. The problem is that no one was required to produce one.

Why This Is Governance, Not Strategy

Compounding the issue, Osterman estimates that “disposable workers,” meaning contractors, freelancers, and gig labor, now make up about 35% of the American workforce. As corroboration he points to the Bureau of Labor Statistics 2023 count of 6.9 million contingent workers, 4.3% of the workforce, up from 3.8% in 2017. Treat the 35% as his estimate rather than a settled figure. The direction it points to still matters: a growing share of output is performed by people who are easy to cut and whose productivity nobody tracks closely either.

Stack those two facts and the shape of the failure becomes clear. A workforce increasingly invisible to measurement, plus a technology whose productivity contribution is asserted rather than instrumented, equals decisions made on vibes at the exact moment vibes are most expensive. Cutting 1,000 jobs is irreversible in any time frame that matters. You cannot A/B test it. You cannot roll it back next quarter if the AI productivity turns out to be thinner than the slide deck promised.

This is why I read it as a governance failure rather than a strategy debate. Strategy is about choosing direction under uncertainty, and reasonable leaders will disagree about AI’s pace. Governance is about the controls that make a consequential, irreversible decision defensible after the fact. The control that is missing here is simple to state: if you cut headcount on the premise that AI replaced the output, you owe a measurement of that output. Same scoreboard, humans and AI, before the decision, not a press-release sentence after it.

We argued a version of this in the two-percent problem, where the realized productivity from AI keeps coming in far below the headline. We traced the workforce inflection in tokenmaxxing and the structural flattening mandate in Meta’s AI-native restructuring. Those threads converge at one point: the moment an unverified productivity number gets cashed in for an irreversible cut. AI washing is the name for cashing it in without an audit.

Do This Now

If you are a CTO, a CFO, or a board member, run one test before your company makes its own AI-justified cut. Ask for the measurement file. Three questions, in order:

1. What was the baseline output of the roles being cut, in units the business already tracks? Tickets closed, deals advanced, releases shipped, cases resolved. If there is no baseline, there is no claim, only a narrative.

2. What is the AI-augmented output that is supposed to replace it, measured on the same units? Not a vendor benchmark, not a demo. Your work, your tools, your people, instrumented.

3. Who verified the delta, and would it survive an auditor reading it back to the board? If the honest answer is “the CEO’s conviction,” you are doing AI washing whether you meant to or not.

A currency-driven layoff can show its math. An AI-driven one should be held to the same standard. The cost of skipping it is not abstract: it is a thousand people, gone, on a number no one could produce.

---

This analysis synthesizes Israeli website creator Wix.com cuts 1,000 jobs, citing strong shekel and AI (Reuters, May 2026), AI part of another tech layoff as Wix CEO announces 20% workforce cut (CNBC, May 2026).

Victorino Group helps boards and CTOs put humans and AI on one measured scoreboard before irreversible decisions get made on unaudited claims. Let’s talk.