If you exclusively read panic-driven headlines about artificial intelligence (AI), then you may assume that AI will inevitably trigger an era of mass unemployment.
Yet a glance at history tells a more nuanced story. Fears of automation displacing workers are nothing new: The Luddites smashed looms in the 1810s; economists warned of “technological unemployment” in the 1930s; and robots were supposed to hollow out manufacturing in the 1980s. Each wave of alarm proved overstated. Recent data suggest the same may be true today: Rather than a surplus of idle hands, we are facing a structural shortage of talent in critical, growing fields as the economy evolves.
An analysis by the Yale Budget Lab found no significant overall impact from AI on jobs since ChatGPT’s debut in 2022. Goldman Sachs Research estimates that even if current AI use cases were expanded across the entire economy, just 2.5 percent of US employment would face displacement risk. Meanwhile, AI-adjacent sectors are already creating new opportunities: Data center construction alone generated an estimated 119,900 direct US jobs in 2024. This surge in demand is likely to persist.
How one interprets these labor market shifts carries both popular and policy ramifications. Americans should rightfully have very different responses to a headline that reads “The End of Work” than to one that instead states “Labor Market Undergoing Latest in Long Series of Changes.” The former justifies hurried, untested policy interventions. The latter warrants studying how to prepare Americans today for the jobs of the future.
Distractions from doomsday scenarios will impede the real work to be done and don’t reflect what is a more complicated issue. Many aspects of AI’s development and deployment will have a positive labor impact beyond just the tech sector. The massive build-out of AI infrastructure, particularly data centers and the energy grids required to power them, demands a surge of specialized engineers, electricians, and tradespeople. The broader construction industry will need an estimated 400,000 more workers by 2033 just to meet baseline demand. What’s more, AI start-ups are creating products and services that will spark new jobs that demand new skills for more people. LinkedIn has already counted 1.3 million new AI jobs, such as data annotator.
In general, however, many workers may need retraining or upskilling in a variety of private and public roles to thrive in the AI age. Designing a comprehensive plan for that transition is beyond the scope of this article. Yet a natural starting point is allowing workers to more easily demonstrate their skills and, by extension, easing the ability of employers to fill positions and improve labor-matching by moving beyond credentialism.
As of now, the trouble connecting workers to employers is made worse by a broken signaling system that traps experienced, talented workers in stagnant roles and locks individuals lacking formal credentials out of growing fields. For decades, employers have relied on four-year degrees as a blunt, expensive proxy for competence, leading to severe degree inflation. Skills-first hiring receives applause on corporate panels but remains incredibly hard to execute at scale. A 2025 survey by Workday, a leading HR technology platform, of 2,300 business leaders found that while 81 percent agreed that skills-based strategies drive economic growth, just 32 percent were confident that their organization actually has the skills it needs; fewer than half reported having the adequate tools to measure competencies at scale. Individual firms cannot justify the steep costs of building high-quality, validated assessments. Because employer-specific tests lack a shared benchmark, results are not comparable, and the signal is simply not trusted. Small businesses, in particular, can’t justify hiring specialists to validate assessments, manage compliance, and constantly refresh test content to ensure applicants are up for the job at issue. Everyone loses in this dynamic: Individuals chase imprecise and costly signals, and employers cannot easily find the right talent.
Policymakers and businesses appear to recognize the advantage of moving beyond credentialism. The House passed H.R. 5235, the Skills-Based Federal Contracting Act, in an aim to reduce arbitrary degree requirements in federal contracting. That’s a welcome move against the so-called “paper ceiling” of degree-based job requirements. But removing a regulatory barrier does not automatically give employers the mechanisms they need to verify competence at scale. Without them, skills-first hiring will remain an empty corporate slogan rather than a functioning labor market. It’s one thing to no longer require degrees for certain roles; it’s another to provide employers with the information necessary to determine that someone is qualified.
The remedy lies in harnessing AI itself to fix the broken signal. AI-assisted assessments can evaluate structured work samples, such as how a candidate diagnoses a fault, executes a playbook, or solves a problem, producing objective, verifiable evidence of competence that no diploma can replicate. This shifts the foundation of labor matching from credentialism to capability. What the market still lacks is a shared benchmark that makes these assessments comparable and trustworthy across firms.
Far from reinforcing stereotypes and presumptions, AI-assisted labor matching helps businesses and workers of all sizes and backgrounds see what their skills truly are and find the best fit. AI lowers the transaction costs of skills-first hiring by establishing unified validation protocols and open datasets and can help eliminate biases about how those skills were learned. For example, a woman reentering the workforce after time as a stay-at-home mom could highlight the time management or event planning skills built through volunteer roles and community organizing. A veteran entering the civilian workforce can translate military experience in leadership, logistics, and operational planning into credentials that civilian employers can actually evaluate. In an economy where AI lowers the cost of learning, traditional degrees should no longer serve as the exclusive gatekeepers to professional opportunity. Ultimately, this framework empowers individuals to pivot based on verifiable capabilities rather than institutional pedigree.
Across multiple sectors, AI is poised to act as a powerful complement to the workforce, not its replacement. AI’s disruption in the labor force may therefore prove to be its own remedy: The same technology that fuels displacement anxieties is precisely what can dismantle the credentialist barriers that have long kept capable workers on the sidelines. The real crisis is not a shortage of work but a shortage of mechanisms that connect the right people to the right opportunities. When capability replaces pedigree as the currency of the labor market, the pool of opportunity does not merely shift, it grows, opening new pathways for workers at every stage of life and from every background.
Wendy Zhang provided excellent research assistance for this article.
