Rethinking Workforce Policy for the AI Era

In April 2026, the House Education and Workforce Committee passed the A Stronger Workforce for America Act (ASWA 2026) on a party-line vote. The bill extended WIOA's authorization through 2032, restructured some service delivery requirements, and introduced a 50 percent training mandate for Adult and Dislocated Worker funds. What it did not do was substantively address the labor market challenge that the legislation will govern for the next six years: the accelerating displacement of workers by AI systems across occupational categories that previous automation waves left largely untouched.

The Senate's 2024 discussion draft had included explicit provisions on AI and workforce transition, recognizing automation as a structural driver of displacement that the WIOA system needed to be equipped to address. Those provisions did not survive into the final House bill. Their absence is not a procedural oversight. It reflects a policy establishment that is still processing AI as an emerging concern rather than a present condition. The World Economic Forum's 2025 Future of Jobs Report found that 39 percent of existing skill sets will be transformed or outdated by 2030, and that AI and automation are the primary drivers. The workforce policy framework that will govern the country's response to that transition was just extended through 2032, largely unchanged.

What Makes AI Displacement Structurally Different

Every major technological transition produces workforce disruption. What distinguishes the current AI transition from prior automation waves is not the scale of displacement, but its location in the occupational distribution.

Previous automation waves, from mechanization through the computerization of the 1980s and 1990s, predominantly affected routine manual and cognitive tasks concentrated in middle-skill occupations: assembly line work, data entry, bookkeeping, and clerical coordination. The labor market response, imperfect and inequitable as it was, involved a set of institutions—community colleges, workforce boards, and WIOA predecessors—that had been built to serve displaced manufacturing and clerical workers.

Generative AI is disrupting occupations at higher levels of the credential and wage distribution. McKinsey's analysis identifies legal research, financial analysis, software documentation, and customer service management among the most exposed functions. These are not occupations served by the existing workforce development infrastructure. A displaced manufacturing worker can access WIOA-funded training at an American Job Center. A displaced financial analyst or legal researcher has no comparable institutional pathway. The WIOA system was not designed for workers at this level of the labor market, and it is not funded to serve them at the scale that AI-driven white-collar displacement will require.

What the Current Policy Map Cannot Handle

Mapping the existing workforce policy infrastructure against the AI-era displacement challenge reveals gaps that are structural rather than incremental.

The planning horizon problem is fundamental. WIOA requires four-year state strategic plans. Generative AI capabilities have changed substantially in less than two years. State workforce plans written in 2024 do not contain meaningful provisions for the occupational categories that AI is currently automating, because the scale of that automation was not visible at the time of writing. A policy framework that cannot update its strategic priorities faster than every four years is not equipped to respond to a technological transition that is measurably reshaping the labor market every twelve months.

The coverage problem is equally serious. WIOA Title I Adult services primarily serve workers who are already unemployed or at risk of unemployment, as measured by income eligibility and other criteria designed to target the most economically vulnerable. Workers who are employed but whose roles are being automated, who have months or years before displacement becomes acute, are largely outside the system's service model. By the time those workers qualify for WIOA services, they have often experienced the full economic cost of displacement rather than accessing retraining early enough to prevent it.

The sector coverage problem compounds the other two. American Job Centers and their affiliated training providers are concentrated in the sectors of construction, healthcare, manufacturing, and transportation that have historically driven WIOA service demand. They do not have established training partnerships or approved provider networks for the financial services, legal, administrative, and technology sectors where AI displacement is most active. Building those partnerships takes time and investment that the current system has not been directed to make.

What Other Countries Are Building

The countries with the most coherent policy responses to AI-era workforce transition share a common feature: they did not wait for displacement to become acute before building the infrastructure to address it.

Singapore's SkillsFuture system has been operational since 2015 and has been iteratively updated, most recently with the Level-Up Programme additions in 2024, to reflect changing labor market conditions. The system provides proactive retraining access to all workers, not just those who are displaced, and includes a national artificial intelligence strategy that explicitly connects AI adoption to workforce transition support.

Denmark's active labor market policy, which has produced the shortest average unemployment spells in the OECD, combines generous unemployment benefits with a strong obligation and infrastructure for retraining. Workers who lose jobs have immediate access to publicly subsidized training, and the system is funded at a level, roughly 2 percent of GDP, that reflects a political commitment to treating workforce transition as a shared social responsibility.

Canada's Future Skills Centre, established in 2018 and expanded under subsequent governments, funds applied research on emerging workforce interventions and provides a testing infrastructure for new policy approaches before they are adopted at national scale. It is specifically designed to address the gap between evidence-generating and policy-implementing timelines that allows disruptive labor market changes to outpace institutional response.

What the U.S. Needs to Build

A workforce policy framework designed for the AI era would differ from the current system in four fundamental ways.

It would plan on a shorter cycle. Four-year state plans should be replaced by rolling two-year plans with annual update requirements, tied to real-time labor market data systems rather than lagging economic surveys. The Department of Labor's investment in workforce data transparency through its performance dashboard is a step in this direction. The planning architecture needs to follow.

It would serve employed workers proactively. The eligibility thresholds that limit WIOA services to the unemployed or near-unemployed should be supplemented by a separate, adequately funded service stream for employed workers in high-automation-risk occupations, accessible before displacement rather than after. Pre-displacement retraining produces better earnings outcomes than post-displacement retraining at every documented point in the research literature.

It would invest in new sector partnerships. Federal workforce funding should include dedicated streams for building American Job Center capacity and approved training provider networks in financial services, legal and administrative services, and technology support roles: the sectors where AI displacement is most concentrated and where existing workforce infrastructure is thinnest.

It would finance continuous learning rather than episodic training. The Individual Learning Account proposals that have circulated in Congress for more than a decade deserve serious legislative attention in the context of AI-driven displacement, where the skills transition is not a one-time event but an ongoing process that will require workers to update their capabilities multiple times across a career.

The choice is not between acting now and acting later. The displacement is already underway. The choice is between building infrastructure proactively, while the window for doing so is still open, and responding reactively, after the scale of disruption has exceeded the capacity of any emergency system to address. The countries that have made the proactive choice are building it. The United States has the 2032 version of a 2014 law.

This is the final article in the PPV Policy Reviews series. Previous policy analyses examined WIOA at Ten, Skills-Based Hiring, State Apprenticeship Expansion, and Credential Inflation. The next and concluding article in the PPV series, Education Systems Built for Jobs That No Longer Exist, examines why the institutional design of education reflects a labor market that has fundamentally changed.

Have research, policy experience, or international comparisons relevant to AI-era workforce policy? Reach out.