AI Rebounding: The Workers Companies Let Go Now Know Their Price
An SHL survey finds nearly 1 in 5 US workers replaced by AI were later asked to return, and 51% of them want a substantial raise to come back.
Key Takeways
- An SHL survey of more than 2,000 US and UK workers found that nearly 1 in 5 US workers say they were replaced by AI and later asked to return, a pattern SHL calls AI rebounding.
- Among workers asked back, 51% want a substantial raise and 42% want a guarantee of job security.
- Gartner predicts that by 2029, 30% of employees laid off due to replacement by AI will need to be rehired, often at considerably higher cost.
- When every competitor rents the same AI, the judgment held by experienced people becomes the advantage that is hardest to copy.
Bring Alex to your organization: keynotes, workshops and advisory, in person or virtual. Check Alex's availability
Nearly 1 in 5 US workers say they were replaced by AI and later asked to come back. The finding comes from a survey of more than 2,000 employed adults in the US and UK, conducted by Propeller Insights for SHL and reported this month by HR Dive and Fortune, and in the UK the share was 13%. SHL calls the pattern AI rebounding, a label that captures this year's workforce decisions with uncomfortable precision.
Think about the person who gets that call, because the numbers start with them. Some months earlier, a manager explained that software would now handle their work, and they carried that news home to the kitchen table and started looking for whatever came next. Now the same employer wants to know whether they would consider returning, and relief is only part of what they feel. The other part is a harder question about what they are worth to people who already decided once that they could be replaced.
The survey is self-reported, so it describes these experiences as workers remember them and should be read on those terms. Even so, the conditions those workers would attach to a return are specific enough to take seriously.
| What workers asked back want | Share |
|---|---|
| A substantial raise | 51% |
| A guarantee of job security | 42% |
| Full transparency on how AI is used | About 40% |
Half of returnees want a substantial raise
Half of them want a substantial raise, a figure worth studying in any finance meeting where a headcount reduction was approved on the assumption that AI would carry the load. Lucy Beaumont, global SVP of product and design at SHL, told HR Dive that AI rebounding “should be a wake-up call,” and she described the behavior behind it in plain terms.
“Organizations are making fast, sometimes myopic, decisions about their people.” Lucy Beaumont, SHL, in HR Dive
Speaking to Fortune, Beaumont described where those decisions can lead, with the business “rehiring the same experts they just let go of, at a far greater cost.” Gartner expects the same arc at larger scale, predicting in a September forecast that by 2029, 30% of employees laid off due to replacement by AI will need to be rehired, often at considerably higher cost.
The cuts behind those forecasts are substantial. Challenger, Gray & Christmas has counted 121,401 announced job cuts in 2026 in which employers cited AI, about 20% of all cuts so far this year, and the monthly pattern is broken down in this look at Challenger's October figures. The longer trend is tracked in our AI job loss statistics. Behind each of those numbers is a household that had a conversation it never planned to have.
Rented intelligence cancels out
When a company cuts experienced people to fund AI, it trades something it owns for something it rents. The model it licenses is available to its competitors on similar terms, so the capability raises every competitor's floor at roughly the same moment. When both sides of a market rent the same intelligence, the rented part cancels out. What remains is whatever each company actually holds, mostly the accumulated judgment of people who have seen thousands of unusual cases and can tell when a model's confident answer is wrong.
Before long, everyone will have the same AI. Nobody will have the same people.
Read that way, the rebound data looks like early evidence of a larger pattern, in which some organizations that cut experienced staff are discovering that the knowledge walked out with them and that the most direct way to recover it is to ask. The returnees understand this, which is why half want a raise and 4 in 10 want full transparency on how AI is used before they agree to anything. Those are the terms of people who know they hold the advantage that is hardest to copy.
History has a rhyme for this. In March 2007, Circuit City let go of 3,400 of its highest-paid, most experienced salespeople to cut costs. By November 2008 the company was in Chapter 11, for many reasons beyond that single decision, and Best Buy had hired a number of the people Circuit City let go. The savings did not rescue the retailer, and some of the expertise went to work for the competitor that survived.
Judgment is the asset worth keeping
I spent 20 years inside a large technology company watching expensive tools arrive with great fanfare, and the lesson I carried out is that a tool is only as good as the person who knows when to overrule it. That person works Above the Algorithm, supplying the judgment and accountability that no license fee covers. Cutting those people to pay for software spends the asset a company owns outright to buy capability its competitors will shortly have as well. The wiser investment is in the people who check the work, a question explored further in this piece on who checks the work after AI job cuts.
The phone call is a negotiation
If you are the one who gets the call, you have more standing than the original layoff made it feel. Before you answer, write down what broke or slowed after you left, because the company asking you back already knows and you should be able to name it as well. Then ask what the AI now does in your role and who reviews its output, since the answer tells you whether you are returning as the expert or as the cleanup crew. If security matters to you, ask for it in writing, which is exactly the guarantee 42% of returnees in the survey said they want.
If you are the leader making that call, the cheaper question arrives earlier. Before the next headcount plan reaches your desk, ask which roles on the list hold judgment the model cannot check for itself, and what it would cost to buy that judgment back in 2 years. Gartner's forecast estimates how often that repurchase happens, and your own spreadsheet can put a price on it. Every rebound call is also a family deciding whether to trust an employer a second time, and a community learning whether experience still earns a living.
The better call comes first
The SHL survey began with a phone call asking people to come back. The better call is the one a leadership team makes before anyone leaves. Take one question into your next leadership meeting: if we cut this role and rent the AI, who here will still know when the AI is wrong? If the answer is nobody, you have found the people worth keeping. And if you would like to think it through privately first, I'm easy to find.
Does AI rebounding mean the original layoffs were a mistake?
AI rebounding is evidence of a pattern across employers, and the SHL survey cannot say whether any single company's cut was wrong, since it records what workers report rather than what companies decided internally. A rehire can follow a change in demand or a discovery that the model needed more human review than expected. The more useful test for leaders is to identify, before any cut, which roles carry judgment that would be expensive to buy back.
How should a returning worker think about pricing a rehire offer?
A returning worker pricing a rehire offer can start from the cost the employer carried after the role moved to AI, including rework and slower turnaround, and from the time a new hire would need to reach the same level. Gartner's forecast that rehiring often comes at considerably higher cost suggests the market already expects a premium. Asking for terms in writing, including how AI will be used in the role, protects the agreement after the urgency fades.
Here is what makes Alex a credible voice on this topic: Alex Goryachev, the former Managing Director of Innovation Strategy and Head of Global Innovation Centers at Cisco, shaped a $1.1B innovation portfolio over 20 years and saw firsthand how decisions about new tools and the people who use them determine which investments pay off.
.svg)
