Artificial intelligence (AI) and related technologies are rapidly disrupting how people work and may fundamentally alter labor market opportunities. AI will change how some jobs are performed, eliminate others, and add new jobs. And the broader economy will evolve as AI impacts productivity and the distribution of income.
Despite vast uncertainty about the eventual impacts of AI, policymakers need to begin working today to ensure the economic gains translate into benefits for workers, businesses, and communities alike. The decisions policymakers make now will shape whether AI expands or narrows economic opportunity for all workers.
To help policymakers meet this pivotal moment, we’ve published a blueprint report for building the inclusive AI economy that workers want and deserve—and for building the evidence policymakers need to make that economy a reality. It reflects our vision for an inclusive AI economy that has four main elements:
- workers have the skills they need to thrive as AI changes jobs,
- workers have high-quality jobs in the AI-powered workplace,
- workers are economically secure during the AI transition and beyond, and
- innovation in AI empowers workers and drives inclusive growth.
In this article, we explain how policymakers at the federal, state, and local levels can help advance these goals. With the right data and evidence, they can create the conditions so that all workers can navigate change and have high-quality jobs, businesses can thrive, and the economy can grow.
Why should policymakers act now?
Policymakers have an opportunity not just to respond to AI’s impacts on workers but also to actively shape the direction of this technological transition. Without thoughtful and data-driven policy, AI could generate productivity gains that never reach most workers, leaving many people and places behind. Research points to four risks that policymakers need to address:
- AI innovation may favor automation at the expense of workers and the public good.
- Workers may shoulder the costs of AI-driven labor market transitions.
- Outdated labor market policies (e.g., wage and hour, privacy, and antidiscrimination protections), may fail to safeguard workers in an AI-enabled labor market.
- Economic benefits from AI will not necessarily be shared with or among all workers.
Below, we identify and describe how policymakers can have the most impact in each of these critical issue areas. The blueprint report offers additional details, including key insights from available evidence, pressing questions, and promising directions for policy reforms and practice innovation.
To help workers get the skills they need to succeed in an AI-powered economy, policymakers should make necessary investments in education and training
AI is already changing the skill sets employers are looking for as they hire and promote workers. To meet the challenges created by a changing economy, education and training policies need to be nimble, responsive, worker-centered, and employer-driven. For education and training policies to be successful, we also need effective tools that help workers find their best-fit jobs—and help employers find their best-fit workers.
Federal, state, and local policymakers have a role to play in helping upskill workers, connecting them with good jobs, and matching employers with workers who have the skills they need.
Help workers obtain and use new skills. Workers who cannot access the skills that AI-era jobs demand are at risk of being left behind—not necessarily because the jobs aren’t there, but because the pathways to building needed skills can be unclear or unaffordable. At the same time, education and training systems are being asked to prepare workers for a future that is still taking shape.
Getting this right requires near-term responsiveness from higher education and training institutions to help workers adapt to jobs that are changing now. It also requires longer-term systems change in how we define, teach, and credential the skills that will matter most, including new and expanded forms of support for on-the-job training and upskilling, such as apprenticeships.
Policymakers can support the education, training, and workforce development institutions on the frontlines of this transition by rethinking policy goals, building integrated systems rather than siloed programs, investing resources at the scale the challenge demands, and prioritizing equitable distribution of those resources.
Use AI to match workers with jobs and employers with talent. AI is transforming how workers search for jobs and how employers screen and hire, and it could support workers’ economic mobility and ease transitions between jobs. But it also risks replicating and further entrenching existing barriers and disparities in hiring and employment. The design and implementation of AI-powered job search and matching tools will have significant implications.
To help workers find good jobs and to support employers with recruiting and retention, policymakers have a role in funding, regulating, and setting standards for AI job search and matching tools—with particular attention to whether they’re widening or narrowing opportunity gaps.
To ensure workers have high-quality jobs in an AI-powered workplace, policymakers should modernize labor market regulations and strengthen worker voice
The quality of work—not just its availability—is at stake in the AI transition. AI could transform low-quality jobs into well-compensated, meaningful work that drives productivity gains for businesses. It could also degrade job quality, prioritizing short-term profits over workers’ well-being. How employers implement AI, how workers and their representatives shape those choices, and how governments set the rules for employers and workers will determine whether AI enhances or erodes pay, autonomy, and other fundamental elements of job quality.
Update policies and regulations to reflect AI’s role in shaping labor markets. Labor market regulations and worker protections, which establish both how labor markets operate and what working conditions are like on a job, are powerful determinants of job quality. But many of these regulations are outdated and struggle to respond to the evolution of modern labor practices. The advent of AI in hiring, employment, and management practices is further testing these limits.
Without updated rules to reflect the role of AI, even well-intentioned businesses will operate in a system that may reward employers at the expense of workers, and workers will have limited recourse when things go wrong. Policymakers at the federal, state, and local levels can develop updated labor market policies and regulations that are responsive to AI, such as by setting new rules and rights around the collection, use, and ownership of worker data.
Incorporate worker voice in AI design and implementation. Worker voice, defined as opportunities for workers to provide input on working conditions, is associated with not only improved job quality, but also higher levels of productivity and innovation. Including workers in decisions about how firms adopt AI will be critical for a technological transition that preserves and enhances job quality while maximizing productivity and minimizing workforce disruptions.
Collective voice and bargaining, including unions, provide promising avenues for including worker voice in the AI transition. But while unions have a key role to play, addressing the challenges posed by AI will also require new and nontraditional forms of worker engagement and collective bargaining, such as forms of sectoral bargaining and coregulation. Policymakers can modernize the rules governing worker voice and power that have not kept pace with technological change and the growing power imbalance between workers and their employers.
To promote the economic security of workers through the AI transition and beyond, policymakers should prioritize supporting displaced workers and sharing prosperity
Past economic shocks driven by changing patterns of trade and advancements in technology have demonstrated the limits of current policies designed to support displaced workers. Social insurance programs like unemployment insurance (UI) provide a floor but have failed to deliver real income security for too many workers. Policies such as Trade Adjustment Assistance (TAA) helped some dislocated workers, but did not operate at the scale or efficiency needed. The longer-term consequences of economic dislocation have been higher rates of joblessness, lower lifetime earnings, and continued decline of already-struggling communities.
Improve policies and programs for displaced workers. Economic security for workers is the foundation of a healthy economy, and job loss due to technological advances such as AI—in the absence of adequate worker supports—can cause workers significant losses in income, both in the near and long term.
Current programs such as UI do not adequately support workers through transitions today, providing only limited support to too few workers. And to help workers remain economically secure as AI reshapes the labor market, programs built for a different era may need to be fundamentally reimagined, not just tweaked. Policymakers can start by strengthening existing programs such as UI, closing the gaps that left too many workers behind in past transitions. But they should also be considering new approaches—such as wage insurance, or new forms of adjustment assistance—that will need to be built from the ground up if they are to provide meaningful income security to all the workers AI will impact.
Ensure gains in productivity and prosperity are widely shared. Productivity growth is the basis of rising living standards, but workers have not shared fully in the gains from rising productivity over the majority of the last half century. Early evidence suggests that productivity gains from AI could flow disproportionately to business owners and shareholders, rather than workers, with the potential to accelerate rising earnings and income inequality. Under some AI scenarios, the economy could grow while workers’ earnings decline and job opportunities diminish.
Policymakers will need to develop corresponding policy responses to ensure gains in productivity and prosperity from AI reach workers, and are equitably distributed by income, education level, race, and place. Doing this effectively will require a range of policy levers, including labor market regulation, tax reforms, updated income support policies, and potentially including new forms of policies such as basic income and wealth-building programs.
To ensure AI benefits workers and drives economic growth, policymakers should use innovation and technology policy levers to guide the direction of AI innovation
The trajectory of AI development is not fixed—but right now, it is being shaped by a relatively small number of private actors, optimizing for a relatively narrow set of goals. The incentives driving AI development do not automatically point toward worker benefit, or toward broadly shared productivity growth. At the same time, underinvestment in publicly supported research institutions have left university-based research labs less equipped to pursue directions of innovation that could benefit society more widely—directions which private actors may not have incentives to pursue.
Use innovation policy to incentivize the development of AI that empowers workers. Innovation policy—including research funding, government spending, data and intellectual property rights, and technology and consumer regulations—shapes what gets built and who benefits. Current incentives may be tilting AI development toward automation and profit rather than augmentation and broadly shared growth. Without deliberate policy action, that default will hold.
Policymakers working across science and technology, tax, and industrial policy as well as firm and consumer regulation, can reorient these incentives. Governments can help shift innovation in workers’ favor by investing in forms of AI that empower workers and by updating the rules and tax incentives that guide how private actors develop and deploy technologies.
Reinvigorate public investment in higher education and research. Some of the most broadly shared economic advances in American history trace back to publicly-funded universities and research institutions—not because government innovates better than the private sector, but because public investments can pursue social goals that markets won’t prioritize on their own.
Policymakers can reinvigorate this model, working with universities to develop AI tools oriented toward the public good, equipping the next generation of workers and innovators with the skills to shape the technology, and helping ensure the next generation of AI builders and technologists reflects a diverse set of voices and goals in pursuit of broadly beneficial economic outcomes.
To ensure AI policy innovation and reforms are grounded in evidence, policymakers should invest in public data infrastructure
One reason for the uncertainty around AI adoption and its effects on the economy and labor market relates to the limitations of existing data. While federal statistical agencies have undertaken innovative efforts to measure AI adoption and untangle its impacts in key data series, they have also been under increasing strain. Private data sources have been important for learning about AI’s early impacts, but these data are typically proprietary. While private data can complement public data, they cannot replace public data as an essential public good.
Policymakers can address this issue by investing in public data infrastructure to support the collection of richer, higher-frequency data, covering key aspects of AI’s use and impacts that current data and statistics struggle to capture. Enhanced data—through improved survey measures, innovations in the use of administrative data, and collaborations with employers, platforms, and providers—would provide the evidence needed to clarify AI’s impacts and develop policies targeting the right challenges and the right workers.
A call for collaborative action
AI could generate widely shared economic prosperity. But without deliberate, evidence-informed action, there are no guarantees that everyone will share in its benefits. And there are many reasons—including the economic outcomes of past technological revolutions—to expect that many workers and communities could be left behind in the AI transition.
But this time can be different. Policymakers can learn from the past and work toward shared prosperity, but it requires action. The economic transformation expected from AI is still in its early days, and the choices that policymakers and other stakeholders make now can bring us closer to an inclusive AI economy that benefits workers, businesses, and communities.
