Artificial intelligence is generating enormous productivity gains for corporations, but the workers whose jobs are being reshaped or eliminated are not sharing in the windfall. Real wages have fallen across wealthy nations during the cost-of-living crisis even as corporate profits outpaced labor costs, and a growing body of research warns that AI adoption will accelerate that divergence. The result is a pressure cooker. If the gap between executive enrichment and worker hardship keeps widening, organized backlash against the companies driving the shift may be unavoidable.
Profits Rise While Paychecks Shrink
The economic backdrop for AI-driven disruption is already hostile to workers. Across OECD member states, detailed analysis of recent labor-market trends shows that real wages fell during the cost-of-living crisis while profits rose faster than labor costs. The labor share of national income, the portion that goes to workers rather than capital owners, has been declining in many of those same countries. That means even before generative AI tools began reshaping white-collar work, the basic bargain between employers and employees was already fraying.
This squeeze is not abstract. When purchasing power drops and corporate margins expand simultaneously, the gap shows up in household budgets: higher grocery bills, deferred medical care, and growing reliance on credit. The pattern suggests that firms have been able to pass rising costs onto consumers and absorb the surplus as profit rather than channeling it into compensation. AI now threatens to intensify that dynamic by letting companies do more with fewer people, further tilting the balance toward shareholders and heightening the sense among workers that the system is rigged against them.
AI Productivity Gains Flow Unevenly
The efficiency case for AI is well documented. A staggered rollout of generative tools at a Fortune 500 software company, studied across 5,179 customer-support agents, found that AI-assisted workers were roughly 14% more productive on average, with the largest gains among novice employees who could lean on the system for guidance and suggested responses. A separate randomized experiment spanning 66 firms and 7,137 knowledge workers showed that generative AI reduced time spent on email, improved the quality of written output, and cut after-hours work, signaling real gains in both efficiency and work-life balance for those who remained employed in these roles.
The trouble is who captures the value. According to PwC’s 2025 Global AI Jobs Barometer, industries most exposed to automation are experiencing a fourfold increase in productivity growth, and workers with advanced AI skills command a wage premium of more than half again what comparable workers earn. That premium sounds like good news until you consider its flip side: employees who lack AI fluency are falling further behind, often trapped in roles that are both more exposed to automation and less likely to see meaningful pay increases. The productivity windfall is real, but it is concentrating in a narrow band of the workforce and among the investors who own the technologies, leaving everyone else more vulnerable.
Wealth Inequality Grows Even if Wage Gaps Narrow
An IMF working paper published earlier this year, designated Working Paper No. 2025/068, adds a counterintuitive wrinkle to the story. Using macroeconomic modeling and distributional data, the authors conclude that in some scenarios AI adoption can reduce wage inequality by boosting the productivity of lower-skilled workers, narrowing pay gaps between occupations. Customer-service agents armed with AI, for instance, may perform closer to seasoned professionals and capture modest raises as a result. Yet the same analysis finds that wealth inequality can rise at the same time, because higher capital returns and stronger incentives to invest in AI technologies channel the largest gains toward shareholders and asset owners.
This distinction matters for the revolt thesis. Worker anger does not track neatly with statistical measures of wage dispersion; it tracks with the visible, growing distance between executive compensation and median pay, especially when layoffs and restructuring are justified in the name of technological progress. The U.S. Securities and Exchange Commission’s requirement that public companies disclose their CEO-to-median-worker pay ratios, adopted through a rule mandating those comparisons in proxy statements, has made the gap harder to ignore. Advocacy groups have seized on the resulting numbers to highlight firms where executives reap stock-based windfalls tied to AI-driven cost cutting while front-line employees face stagnant wages, unstable schedules, or displacement, deepening the perception that AI is being used as a tool for extraction rather than shared prosperity.
Displacement Scale and Who Bears the Cost
The World Economic Forum’s Future of Jobs Report 2025, based on a survey of more than 1,000 companies representing over 14 million workers, projects significant job disruption through 2030 as employers accelerate automation and reconfigure roles around AI. While the report anticipates new positions in data analysis, machine learning, and AI governance, it also foresees substantial displacement in routine office work, basic customer service, and certain manufacturing tasks. The International Labour Organization’s task-level analysis, which examined roughly 30,000 work activities, similarly concludes that one in four jobs globally will be transformed by generative AI, either through partial automation of tasks or full elimination of roles that can be restructured around software.
The burden will not fall equally. Workers in smaller firms and less affluent regions are less likely to receive structured retraining, and those without savings or social safety nets will struggle most with periods of unemployment or underemployment. Even when new jobs emerge, they may be geographically distant or require credentials that displaced workers do not yet have. Without deliberate policies to support transitions, such as wage insurance, portable benefits, subsidized upskilling, and stronger collective-bargaining rights, the costs of AI-driven restructuring will be borne disproportionately by people who had the least say in adoption decisions, feeding resentment toward both employers and policymakers.
Can Policy and Corporate Choices Avert a Revolt?
Whether AI-triggered backlash materializes as scattered protests or a broader labor revolt will depend heavily on how governments and corporations respond in the next few years. Transparency is one pressure valve: when firms clearly communicate how AI will be deployed, what protections exist against arbitrary layoffs, and how productivity gains will be shared, workers have more reason to see technology as a joint project rather than an existential threat. Tying executive bonuses to metrics such as median-wage growth, internal mobility, and training completion (rather than purely to headcount reductions or short-term earnings) can also align incentives away from using AI primarily as a cost-cutting weapon.
Public policy will shape the playing field on which these choices are made. Governments can encourage more equitable outcomes by attaching labor conditions to tax incentives or public contracts for AI-heavy firms, requiring commitments to retraining and redeployment where feasible. Strengthening data on job transitions and funding independent evaluation of reskilling programs would help identify which interventions actually move displaced workers into comparable or better roles. Finally, updating labor law to clarify how algorithmic management, AI monitoring tools, and automated scheduling intersect with existing protections could give workers firmer ground from which to negotiate. If AI continues to magnify the gap between corporate profits and household security, anger is likely to grow; if its gains are visibly and fairly shared, the promised productivity revolution may yet avoid triggering a revolt.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
