Nearly 1 in 5 US Workers Replaced by AI Say They Were Asked Back

What happened

On 30 October 2026, CNBC reported on a survey commissioned by SHL, a talent acquisition and talent management company. The fieldwork was carried out by the market research firm Propeller Insights, which polled more than 1,000 US workers and a similar number in the UK. According to CNBC, nearly one in five US respondents said they had been laid off or replaced by AI and later asked to return. Lucy Beaumont, SHL's global senior vice president of product and design, said employers were realising they had “pulled the plug too quickly” on their people.

What the returning workers asked for

Most respondents were open to going back. CNBC reports that only 6.6% of US workers said they would refuse to return under any circumstances. That figure covers all US workers surveyed, not only those who had been asked back. Among the US respondents, almost half would require a significant pay rise, 43% wanted written guarantees of job security, and 40% wanted full transparency about how AI would be used in their role.

The numbers vary slightly between reports. A CFO Tech write-up of the same study gives 16% for the combined US and UK sample of 2,036 adults, and a 51% figure for the pay demand. A second write-up gives 42% for written guarantees. Readers who need exact figures should check SHL's own report.

Why it matters for businesses that automate work

The survey does not stand alone. Orgvue's March 2026 report found that 32% of organisations that had cut staff because of AI had rehired. Challenger, Gray & Christmas said employers credited AI as the main reason for nearly 40% of the job cuts announced in May 2026. CNBC's 1 July 2026 report described Ford rehiring more than 350 experienced engineers after automated quality-control systems failed to capture veteran expertise. It also described the Commonwealth Bank of Australia reversing cuts after replacing more than 40 customer service staff with an AI voice bot in 2025.

The lesson for any business is narrow. Automation can run a fixed set of steps quickly. It does not carry the context behind exceptions, customer history or the judgment that long-serving staff build up. Where a role was cut before anyone wrote down what that judgment was, the gap tends to show up later.

What changes in practice

For a five-person firm or a group with thousands of staff, the sequence is the same. Map the work before any role changes. Decide which steps a system will run, which exceptions go to a person, and who checks the output. The table below sets out a simple version.

Be clear with the people involved. The SHL respondents asked for written job-security commitments and a plain account of how AI would be used. A business that writes both down before a change will be in a better position than one that explains them after. Any reduction in roles, pay or hours also triggers the notice, consultation and severance rules that apply in the country concerned, and the change should be agreed with the employee or checked with a lawyer.

What the survey does not show

The SHL figures are reports from workers about invitations they received. They do not show how many people accepted, how many stayed, or how well the returns worked out. The 5% margin of error that one report attaches to the study covers sampling, not the wording of the questions. The research is useful as a signal about worker expectations. It is not a measure of how common rehiring is across all employers.

What to watch next

Three things will show whether this trend is lasting. The first is whether SHL or Propeller Insights publish completed rehire data, rather than invitation data. The second is whether Orgvue's 32% finding is repeated in later surveys. The third is the forecast reported by HCA Mag, which says 30% of employees replaced by AI will be rehired by 2029, linking the view to Gartner. That forecast is a projection, and it should be treated as one.