Nearly every CEO surveyed in Mercer’s 2026 Global Talent Trends report expects AI to drive layoffs at their company within two years. That is not a fringe prediction or a cautionary hypothetical; it is the stated plan of the people running the world’s largest employers.
The Mercer survey found 99% of executives are prepared for AI-driven headcount cuts in the short term. Only 32% believe their workforce can optimally combine human and machine capabilities. Those two figures together are more revealing than either alone: most CEOs see AI replacement, not AI augmentation, as the more likely outcome of the transition they are managing.
READ: Branch Cuts Jobs in Nigeria and Kenya Despite $30M Profit
AI Layoffs Have Already Started - and Entry-Level Workers Are First
The numbers are not theoretical. Challenger, Gray & Christmas tracked over 85,000 tech job cuts through April 2026, up 33% on the same stretch last year. Meta cut 8,000 roles, about 10% of its headcount. Intuit followed with 3,000 cuts, roughly 17% of its global workforce, pointing directly to AI integration as the reason.
The pattern of who gets cut is consistent. Mercer found that most expected headcount reductions will concentrate on early-career positions. AI systems are most capable at the structured, repetitive tasks that companies traditionally assigned to junior staff as on-the-job training. Those entry points into the workforce are closing. Anthropic’s Dario Amodei has put a number on it: up to half of all entry-level white-collar jobs gone within five years.
New York Fed data puts the unemployment rate for recent graduates at 5.6%, a full percentage point above the 35-year average of 4.5%. The job market for 22-to-27-year-olds is, by multiple measures, the worst since the height of the pandemic.
READ: Paystack's New Dashboard Answers Your Questions - Here's What Changed
The Workplace Toll Goes Beyond Job Losses
Only 44% of employees reported thriving at work in 2026, down from 66% in 2024, according to Mercer. The drop tracks directly with anxiety over AI-driven displacement. Researchers are now proposing the term “AI replacement dysfunction” AIRD to describe the existential distress workers report as companies announce automation roadmaps.
The anxiety shows up in how workers engage with the technology itself. Gen Z’s use of AI tools is plateauing, and members of the cohort increasingly report feeling anxious and angry about the technology rather than optimistic. An NBC News poll from March found AI more unpopular among U.S. voters than almost any other issue tested.
The stakes extend beyond individual workers. The jobs that absorbed junior developers, data entry analysts, and customer support specialists in Lagos, in London, and in Manila are precisely the roles AI is most aggressively automating first. The Branch layoffs in Nigeria and Kenya, confirmed last month despite $30 million in global profit, are one regional example of a pattern playing out across every market where digital-first companies face investor pressure to cut costs.
READ: Google Gemini Omni Flash Turns Any Input Into Video
The Honest Version of What Comes Next
The optimistic case is real and worth stating. The U.S. economy has still added 304,000 jobs this year. Amazon’s Jeff Bezos told CNBC he expects AI to create a labour shortage, not a surplus. Several CEOs at Web Summit Vancouver said they are hiring more than before, pointing to net growth even as they automate specific functions. Sim Desai, CEO of Hiive Capital, was direct: “In the short term, there’s a lot of job creation because a lot of people are investing in adopting AI tools.”
That is true. It is also incomplete. Job creation at the top of the skills ladder does not automatically absorb workers displaced at the bottom. A junior developer replaced by an AI coding assistant does not simply become a senior AI systems architect. That gap is measured in years of training, access to resources, and institutional support. None of those factors appears in a CEO’s AI ROI calculation.
READ: The Korean Startup Building the Lens Inside AI Glasses
The workers most exposed to this transition have the fewest options when it hits. In markets without unemployment insurance or meaningful retraining infrastructure, the exposure is harder, and the landing is worse. The question governments, universities, and employers need to answer is not if AI will reshape the labour market. It already is. The question is whether the infrastructure to catch displaced workers will exist before the displacement accelerates.
