The Sentence That Buried the Narrative

“We went too far.”

Sebastian Siemiatkowski, Klarna’s CEO, publicly admitted in 2026 that the company overcorrected. Klarna — which in 2024 boasted about cutting from 4,000 to under 3,000 employees and replacing 700 support staff with AI — started rehiring.

When I read this, I thought of all the triumphant 2024 LinkedIn posts: “AI replaced 700 jobs at Klarna! The future is here!” Now the future has a name: regret.

And Klarna isn’t alone. Gartner published on February 3, 2026 a prediction every executive should read: by 2027, 50% of companies that cut headcount attributing it to AI will rehire for similar functions — often under different job titles.

The Regret Numbers

The convergence of data from multiple independent sources makes the pattern impossible to ignore:

55% of employers who made AI-attributed cuts regret it, according to Forrester’s Predictions 2026: The Future of Work report — independently corroborated by Orgvue’s annual workforce study.

50% of AI layoffs will be reversed by 2027, per Gartner (February 2026). Kathy Ross, Gartner Senior Director Analyst: “While AI-driven layoffs have captured attention, the reality is more complex. Most recent reductions were influenced by broader economic conditions rather than automation alone.”

Only 20% of customer service leaders actually reduced staff due to AI, per Gartner’s October 2025 survey of 321 leaders. Most report headcount steady or growing.

57% of executives expect AI to increase headcount next year, vs only 15% expecting reduction (Forrester 2026). The sentiment inversion is nearly total.

Two-thirds of employers who eliminated AI-attributed roles are already rehiring, per Careerminds (February 2026, 600 HR professionals). Over half did so within 6 months. Only 2% waited more than a year.

5.3% is the rehire rate of terminated employees, tracked by Visier — and it’s climbing.

The Pattern: Fire in Headlines, Rehire in Silence

What bothers me most: layoffs were public, rehires are silent.

Klarna press-released replacing 700 employees. When it started rehiring, no announcement. Ford laid off thousands of experienced engineers. When it discovered automation lacked the tacit know-how, it rehired and promoted over 350 engineers, per American Bazaar. No press release.

Amazon cut 16,000 in a single January 2026 week — alongside Nike (775) and Home Depot (800) in the same “Automation Week.” Zuckerberg later admitted he “made mistakes.”

The pattern is consistent: announce AI-attributed cuts (Wall Street applauds, stock pops), discover AI doesn’t know what humans knew (quality drops, customers complain), quietly rehire with different titles to avoid explicitly contradicting the narrative.

As Forbes analyst Jon Markman wrote: “Stocks that pop on layoff announcements are pricing in margin expansion. Separate companies making structural transformations from the ones making announcements.”

The Convenient Lie

The most revealing data point: only 20% of service leaders actually reduced staff because of AI. The other 80% of AI-attributed layoffs were driven by broader economic conditions — pandemic over-hiring, margin compression, operational restructuring.

Saying “we cut because of AI” was socially acceptable. “‘We’re an AI-first company now’ plays a lot better on an earnings call than ‘we over-hired during the pandemic and want to juice margins.’”

Inc. synthesized the root error: “Companies reversing course replaced jobs requiring human judgment with AI that delivered information retrieval.” They confused search with analysis. Information with judgment.

The Cost of Reversal

Rehires aren’t cheap. When you fire an engineer with 10 years of tenure, you lose institutional memory — the “remember when DNS gets wonky, check that system in the corner” I discussed in the Amazon post. You lose client relationships. Context no onboarding recovers.

Robert Half (2026) found 29% of companies that cut for AI already reopened the same positions. US unemployment claims fell to their lowest since 1969 — the market is tight, rehiring costs more than retaining would have.

What I Take from This

“AI will replace jobs” is incomplete. The correct version: “AI will replace tasks — and create demand for new tasks requiring human judgment.” Gartner predicts AI will create more jobs than it destroys starting 2028. WEF estimates a net positive of 78 million positions.

AI is an accelerator, not a replacement. “AI is an incredible accelerator, but a catastrophic failure when used as a complete human replacement.” Use AI to make your team faster. Not smaller.

If you were laid off “because of AI” — there’s a 55% chance it wasn’t because of AI. It was over-hiring, margins, or narrative convenience. That doesn’t diminish the pain. But it contextualizes reality.

Conclusion: The Boomerang Is in the Air

The “Layoff Boomerang” is documented by Gartner, Forrester, Careerminds, Robert Half, Visier, and Forbes. Companies that rushed to replace humans with AI discovered AI doesn’t know what it doesn’t know — and human judgment doesn’t install with pip install.

If you’re a professional impacted, the market is correcting. If you’re an executive considering cuts, read the data first — 55% chance of regret. And if you feel your value is shrinking: the opposite is happening. Your judgment has never been more valuable.

Share if this brought perspective:

55% regret it. 50% will reverse. And the “Layoff Boomerang” is proving what should have been obvious: AI is a tool, not a replacement.

Read Also