AI layoffs save little money but leave workers with lasting scars, data shows
New research shows most companies aren't cutting costs through AI, while workers pushed out by it take longer to find work and often earn less than before.
What happened: Three new reports line up to describe the same mismatch. Gartner found that despite heavy AI spending, 97% of organizations can't cut costs through productivity gains alone, and 38% are already over budget on AI with more spending planned. Separately, a SHL survey found nearly 1 in 5 workers who were laid off or replaced by AI have been asked to come back, and Orgvue found 32% of companies that cut jobs over AI had to rehire. Meanwhile, Goldman Sachs economists tracking workers actually displaced from AI-exposed jobs since 2025 found they take about a month longer to find new work and are 10 points more likely to still be unemployed a year later than other displaced workers.
Why it matters: Many of this year's AI-linked layoffs were made on the assumption that AI could quickly replace labor costs. Gartner's numbers suggest that assumption was premature for nearly all organizations. For workers caught in the gap, the cost is not abstract: Goldman's data shows AI-displaced workers are following the same damaging pattern seen in four decades of past tech disruption, with slower re-employment, weaker pay recovery and compounding effects for workers over 35. Younger workers face a different problem, struggling less once displaced but facing reduced hiring into AI-exposed roles in the first place. Together, the data suggests companies are paying a price for cutting too fast, and workers are paying a larger one for being caught in it.
How it works, plainly: Gartner's research shows AI is freeing up employee time, but only 21% of that freed time gets redirected into higher-value work, meaning organizations capture less value than projected. Goldman can't ask people directly if AI took their job, so it built a score combining an occupation's AI exposure, its industry's AI adoption rate, and whether the job was eliminated, then treated the top 30% as likely AI-displaced. Comparing this group to other job losers since 2023 reveals the gap. SHL's survey of workers who already lived through an AI-related layoff found most would return, but only with conditions: nearly half want more pay, 43% want written job security guarantees, and 40% want clarity on how AI will be used in their role.
The rollout: Some employers, including large companies like Ford and Amazon, have already reversed AI-related cuts. But both researchers and company leaders warn rehiring doesn't undo the damage: trust and morale take a lasting hit, and returning employees tend to show weaker productivity and engagement than before. Workers, meanwhile, are trying to get ahead of the next round by teaching themselves AI tools, building skills that are harder to automate, and leaning into interpersonal and leadership work. Goldman and Gartner both caution it is early: the first wave of AI-displaced workers is still being measured, and leaders are still trying to figure out how to convert AI's freed-up time into something durable instead of just a smaller headcount.