In May 2026, employers named AI as the primary reason for almost 40% of announced job cuts, according to outplacement firm Challenger, Gray & Christmas. Six months on, a measurable share of those decisions is reversing. For CHROs and People & Culture leads, the reversal is not a footnote: it is a line item with a price tag, a trust deficit, and a negotiating counterpart who now holds more leverage than the first time around.

Nearly 1 in 5 US workers surveyed by talent management firm SHL say they were laid off or replaced by AI and then asked to return. Lucy Beaumont, global SVP of product and design at SHL, reads the data as evidence that organizations moved faster than their process knowledge could afford. The rehire conversation is the bill arriving.

How large is the AI reversal trend?

The SHL figure does not stand alone. A March 2026 report by workforce planning company Orgvue found that 32% of organizations that made reductions on the cost-saving promise of AI had to rehire staff because the projected savings did not materialize. Named US employers, including Ford and Amazon, have reversed course after AI-related layoffs in the past year, though neither has disclosed specific rehire volumes.

The SHL research was commissioned from Propeller Insights and covered more than 1,000 US workers and more than 1,000 UK workers between June and July 2026, with a margin of error of 5%. The figures cited in this article refer to US respondents. HR teams benchmarking against this data should weigh SHL's commercial interest as a vendor and look for corroboration from independent sources such as Bureau of Labor Statistics (BLS) Job Openings and Labor Turnover Survey data before treating it as a planning baseline.

For HR leaders who ran AI-driven reductions-in-force in 2025 or early 2026, the WARN Act (Worker Adjustment and Retraining Notification Act, 29 U.S.C. §2102) required 60-day advance written notice to affected employees for qualifying mass layoffs. That statutory paper trail now forms part of the institutional memory a returner brings back to the negotiating table, and employment counsel should be in the room when written guarantees are discussed.

What returning employees now put on the table

The door is still open. Only 6.6% of US workers say they would refuse to return to a former employer under any circumstances. The conditions attached to a return, however, have hardened. Beaumont expects returners to be considerably more demanding of what they expect from an employer. According to SHL, US workers asked back after an AI-driven separation would set the following conditions:

• Almost half would require a significant pay increase.
• 43% would want written job-security guarantees.
• 40% would require full transparency about how AI will be used in their specific role before rejoining.
• 30% name a formal apology or acknowledgment from leadership as a key condition of return.

Taken together, these conditions function less as a pay negotiation and more as a governance audit. Beaumont frames the underlying question every returner is silently posing as one of reassurance: they want to know that the same sequence of events will not repeat. For HR leaders, that question needs a substantive operational answer, not a verbal commitment.

In practice, this reshapes the offer stage. A returning hire carries institutional knowledge and a precise recollection of how the exit was handled. Salary discussions are likely to open at the prior package plus a premium, and contract language around AI governance may surface before the first formal interview concludes. Society for Human Resource Management (SHRM) guidance on structured offboarding and rehire protocols becomes directly relevant here: organizations that documented role scope and AI deployment plans during the layoff phase will have a stronger position when those terms are renegotiated.

The cost that never appears in the rehire budget

Joseph Fuller, professor of management practice at Harvard Business School, describes the logic that drives many of these reductions: to fund a new technology, companies look for savings in the budget, and with previous technology deployments those savings have historically come from laying people off, treating headcount as the most flexible line in the business case.

The rebound exposes the structural flaw in that logic. Fuller argues that the more an individual understands important processes, the more value the organization can extract from its AI investment. Cut the people who hold that process knowledge, and the return the reduction was meant to fund can shrink accordingly. Beaumont makes a parallel point: AI cannot fully replace people who know the business, know its customers, bring judgment, and provide the human connection that client relationships require.

Then there is the trust dimension. Beaumont is direct about what happens to the employment relationship when a worker is let go and later invited back: the organization has broken that trust, and culture is genuinely difficult to change once it has been damaged. That damage does not stay confined to returners. The colleagues who remained watched both the reductions and the reversal. That experience tends to surface in exit interviews, engagement scores, and public employer reviews, with a lag that makes it hard to trace back to the original decision.

For HR and finance teams modeling AI-driven workforce changes, the Department of Labor (DOL) Employment and Training Administration tracks layoff and rehire data through the Mass Layoffs Statistics program and related WARN Act filing disclosures. Cross-referencing a proposed reduction against that data, and against WorldatWork annual Salary Budget Survey benchmarks for returner premiums, gives the business case a more defensible cost structure before it goes to the board.

What HR can do before the next automation decision

Organizations that manage AI-driven workforce changes well tend to treat a headcount reduction linked to automation as a reversible bet, not a one-way saving. Five moves make that bet cheaper to unwind if the savings do not arrive on schedule:

• Map process knowledge before mapping headcount. Identify which roles hold the customer, operational, or judgment knowledge your AI tools depend on, and document it before any reduction list is drafted.
• Price the reversal into the business case. With 32% of organizations in the Orgvue study forced to rehire, model a rehire scenario upfront: recruitment costs, a returner pay premium, and the productivity gap while positions stay open. WorldatWork Salary Budget Survey data provides a benchmark for budgeting that premium.
• Publish how AI will be used in each affected role before the reduction, not after. Returners are demanding this as a condition of reengagement. Offering it proactively to current staff costs far less than negotiating it under pressure six months later.
• Settle your position on job-security commitments before a returner asks. What a written guarantee can cover, and how enforceable it is, depends on the employment framework governing the contract. Under the National Labor Relations Act (NLRA), written commitments that affect terms and conditions of employment can carry collective bargaining implications. Involve employment counsel early. Redeployment or retraining commitments may be easier to honor than open-ended promises, and SHRM's HR competency framework provides a starting structure for documenting those commitments.
• Build internal mobility into the AI roadmap from the start. Beaumont urges leaders to examine what upskilling employees need, what new roles AI creates, and what internal mobility the shift opens up.

SHL found that US respondents are already investing in their own adaptability: 44% are self-teaching AI tools, 42.6% are building industry-specific skills that are harder to automate, 40% are developing interpersonal and leadership skills, and 38% are strengthening creative and strategic thinking. Channeling that self-directed effort into structured reskilling programs, before a separation occurs, is cheaper than buying it back at a premium after a layoff. For organizations with established 401(k) profit-sharing or ERISA-governed benefit structures, tying reskilling investment to those frameworks can strengthen retention without requiring a headcount guarantee.

Planning for sustained workforce disruption

Fuller predicts that AI will create sustained, recurring disruption within many organizations. If that prediction holds, the AI-driven reductions of 2025 and 2026 are a first cycle, and the way organizations unwind them will set the terms of trust for the next one.

The organizations best placed for that cycle will treat employee sentiment as operational data: tracking how reductions and reversals are perceived internally, benchmarking returner compensation against real market salaries through BLS Employment Cost Index data and WorldatWork compensation surveys, and establishing AI transparency standards before candidates demand them at the offer stage. The rehire is the visible cost on the income statement. Eroded trust is the one that compounds quietly, and shows up later in voluntary attrition, employer brand scores, and the difficulty of recruiting the next cohort of process-critical talent. Challenger, Gray & Christmas data and the SHL survey together signal that this cycle is already underway. HR leaders who build the reversal scenario into their AI workforce planning now are buying optionality; those who do not are building a larger bill.