Remember 2025, when every CEO on the keynote circuit swore the future was “AI-first” and fired half their customer service team to prove it? Reality has now sent those CEOs a strongly worded email — from the people they fired, who are currently billing them $200/hour as contractors to clean up the mess.
A new wave of reports this week confirms what frustrated customers have been screaming into the void for a year: the Great AI Layoff was, in large measure, a very expensive mistake.
The numbers are brutal
According to research circulating this week, 55% of companies now regret their AI-driven layoffs. About 29% have already rehired workers they cut in the name of efficiency, per a Robert Half study. More than a third brought back more than half the roles they eliminated. Over half did so within six months — which, in corporate time, is the speed of light.
The most damning stat: 1 in 3 employers spent more on restaffing than they ever saved from the original layoffs. The AI revolution, for these companies, was a round-trip ticket to the same place they started — minus a fortune in severance, recruiting fees, and brand damage.
What actually went wrong
It turns out that “AI can resolve 70% of customer inquiries” is a statistic that works beautifully in a McKinsey slide deck and slightly less beautifully when a real person is trying to dispute a fraudulent charge at 11 p.m. on a Sunday.
Companies fired people for technology those people were never trained to use, based on capabilities that don’t yet exist, and then acted surprised when their Net Promoter Scores cratered and customers started churning to competitors who still answered the phone. E-commerce platforms, fintech apps, and content companies have all quietly reversed course — hiring back writers, engineers, and support reps they paraded out the door eighteen months ago.
The “boomerang” tax
Here is the cruel part for shareholders: rehiring is almost never cheaper. The workers who got boomeranged back typically negotiated higher salaries, better titles, or contractor rates. Some refused to come back at all, leaving companies to pay recruiter fees to fill roles they already had filled in January 2025.
And the institutional knowledge — the thousand small things the 14-year customer service veteran knew about the product that no wiki ever captured — mostly left the building and did not come back. The chatbot, meanwhile, is still there. It’s just now supervised by the human it was supposed to replace.
The emerging consensus
Only about one in five companies surveyed reported that AI fully replaced the eliminated roles without operational issues. The rest are somewhere between “quietly rehiring” and “publicly denying the quiet rehiring.” Meanwhile, management consultants are now selling a new product: “AI transition strategy” consulting — which, as best we can tell, is the service of telling you not to do the thing the same consultants told you to do two years ago.
The job market is adjusting. The workers who got boomeranged back are now sitting on leverage they did not have before. Their LinkedIn profiles read, in effect: “Previously: fired because of AI. Currently: essential because of AI.”
Fire humans. Hire chatbot. Rehire humans to fix chatbot. Hire consultant to explain why. Capitalism is a beautiful loop.