AI layoffs often target roles based on what AI is expected to do, not what it actually does yet. That gap between projection and reality is why so many companies end up rehiring. Gartner expects nearly 30% of employees let go because of AI to be brought back, often at a higher cost than before.

If you're a hiring manager who just approved a headcount cut, or an engineer watching your team shrink, this affects you directly. The pattern shows up across customer service, IT operations and engineering support functions. Roles disappear on a slide deck, then reappear six months later as a job posting, usually with a longer list of requirements and a bigger salary attached.

That's not a reason to panic. It's a reason to look closely at how these decisions get made, and what it costs when they go wrong.

Companies rehire because AI usually automates a task, not a whole job. The parts that get left behind, exception handling, judgment calls, customer trust, still need a person. When that becomes obvious, the company is back in the market for the same skill set it just removed.

This isn't a hypothetical. Gartner's research team has flagged the same risk for customer-service functions specifically, estimating that at least half of companies that cut service headcount because of AI may rehire for similar roles by 2027. Some of that has already started. Reports describe companies bringing back human support staff after automation led to more complaints, not fewer.

There's also a talent-availability problem. The person who got laid off doesn't sit around waiting. They take another job, often at a competitor. When the company realizes it needs the role back, it's not rehiring the same person for the same salary. It's competing for someone new, in a tighter market, often at a premium.

Add onboarding time on top of that. A new hire needs months to rebuild the context the previous person had for free. Most workforce-planning spreadsheets don't even include that cost line, but it's real, and it's often the most expensive part.

The visible cost of an AI-driven layoff is the salary saved. The hidden costs are severance, recruitment fees, onboarding time, lost productivity and the knowledge that walks out the door with the person you let go.

"Rebuilding a cut role typically costs one and a half to two times the salary you 'saved'.", Lee McCabe, Claymore Partners

Research cited by iProspect and Orgvue puts a number on this: companies may spend roughly $1.27 for every $1 saved through workforce reductions, once severance, lost productivity and rehiring costs are factored in. Careerminds data found that nearly a third of organizations, 30.9%, ended up spending more on rehiring than they had originally saved through automation-related cuts.

None of this shows up in the first quarter after a layoff. It shows up a year later, when the support tickets pile up, the senior engineer who understood the legacy system is gone, and the replacement hire needs six months to get to the same level of output. By then, the savings on paper have quietly disappeared.

There's a second cost that's harder to put a number on: reputation. Teams remember who got cut and why. Candidates talk to each other. A company that lays off aggressively and then rehires for the same roles within a year builds a track record that makes the next hiring round slower and more expensive, because good candidates start asking harder questions before they say yes.

No, not in aggregate. AI is shifting which skills are in demand, not eliminating the need for technical talent. A 2026 analysis of euro-area firms found no evidence of broad labor shedding tied to AI adoption. Instead, AI adoption correlated with higher employment expectations, and skill shortages, not AI-driven cuts, were the most common obstacle firms reported.

In the Netherlands specifically, the share of job postings requiring AI-related skills grew from 1.5% in 2022 to 2.1% in 2025, an increase of roughly 5,000 vacancies. That's growth, not contraction. At the same time, ManpowerGroup found that 73% of Dutch employers struggled to fill vacancies in 2026, with AI model development and applied AI among the hardest skills to find.

The tech layoff numbers get attention because they're visible and dramatic. Layoffs.fyi tracked 128,536 tech employees affected across nearly 300 companies globally through early September 2026. What gets less attention is that some of the same companies cutting roles are simultaneously hiring aggressively for AI-specific positions elsewhere. It's not fewer jobs. It's a reshuffling, and the reshuffling is uneven and often badly timed.

Companies avoid layoff regret by treating AI adoption as a redesign of tasks and workflows, not a straight headcount cut. That means testing what AI can actually do before removing the people who currently do it, and keeping a reserve of expertise for the parts that don't automate cleanly.

In practice, that looks like a few concrete habits. Model the full cost of a layoff, including a realistic rehire scenario, before signing off on it. Redeploy people into adjacent roles where the underlying work is changing rather than disappearing. Keep former employees in an alumni network instead of cutting ties completely, because you may need them again within eighteen months, and a warm relationship is cheaper than a cold search.

This is where we've seen the clearest pattern in our own conversations with hiring managers. The teams that get this right don't ask "what can AI replace." They ask "what does this role look like in a year, and who do we need to run it." That's a harder question, and it takes longer to answer, which is exactly why a lot of companies skip it.

A structured intake process forces that question early, before a role gets posted or cut. It's one of the reasons our Delivery Sprint process starts with understanding what the role actually needs to do in twelve months, not just what the job title used to mean.

The roles replacing the ones cut for AI usually demand more, not less. A support engineer now needs to supervise an AI system, not just answer tickets. That combination, domain expertise plus AI-tool oversight, is harder to hire for than either skill alone.

We're seeing this directly in the type of briefs coming in. A role like a senior site reliability engineer today often includes expectations around monitoring AI-driven automation, not just infrastructure uptime. Similarly, teams building out senior AI engineering capacity are competing for a much smaller pool than the general software engineering market, because the hybrid skill set didn't exist in this form three years ago.

The European Commission estimates the EU will need 6.2 to 7 million AI-related workers by 2027, with roughly 60% of the overall workforce needing some level of AI skill. That's not a niche specialization anymore. It's becoming baseline. Companies that cut experienced staff assuming AI would cover the gap are often the same ones now scrambling to hire the hybrid profile that could have prevented the gap in the first place.

This is exactly why quality of match matters more than speed of fill. A rushed hire into a hybrid role, one that needs both technical depth and comfort working alongside AI tooling, tends to fail faster and costs more to replace than a slower, better-matched search.

AI layoffs aren't wrong by definition. Rushed ones are. The pattern is consistent enough now to plan around: cut a role assuming AI covers it fully, discover the gap a year later, then pay more to rebuild what you had.

If you're planning workforce changes around AI adoption, or you're rebuilding a team after realizing the cut went too far, that's exactly the kind of hard-to-fill, hybrid role search we run through our Delivery Sprint process. Roles like a senior enterprise network engineer or a site reliability engineer position across the EU increasingly sit right at that intersection. Worth a conversation before the next round of decisions gets made, not after.

1. 128536 Tech Jobs Lost In 2026 So Far - Times Now
2. AI Layoff Reversals: Every Company That Rehired Workers ...
3. Tech firms scale back AI-driven layoffs as results fall short
4. Tech layoffs 2026: As AI takes center stage, here's a list of ...
5. From Oracle to Amazon: Tech giants drive global wave of layoffs in ...
6. Oracle and others, including Uber, Apple, Microsoft, cut ...

Written by our AI, read by a flesh-and-blood recruiter.