Last fall, I was speaking with the mayor of a midsized city about artificial intelligence. He was worried his team couldn’t predict exactly which jobs were most at risk in his region. I told him he had a bigger problem.

Politicians worry about which jobs will disappear. Economists model unemployment scenarios. Tech executives race ahead with new AI capabilities. Across sectors, leaders are fixated on forecasting disruption — as if knowing what’s coming will be enough to manage the upheaval already underway. It won’t be.

America’s real vulnerability isn’t AI. It’s that we never built a system to help people gain skills, find work, and transition after layoffs. AI is now exposing this failure nationwide.

For more than a century, the field of economic development has evolved in response to major economic transitions — from industrialization, to deindustrialization, to globalization. AI represents the next transformation, but our civic institutions haven’t kept pace. It’s time for economic development and civic leaders to modernize America’s jobs system.

A functioning jobs system brings employers, educators, governments, and people together in clear pathways, so job matching, reskilling, and employment happen quickly, rather than through today’s maze of disconnected programs and institutions.

Without that system, workers face longer unemployment and fewer pathways to stable jobs. Employers face hiring delays and skill shortages. Communities feel the damage through shuttered businesses, weaker tax bases, and public service cuts.

Most of our civic institutions were designed for an older world and never adapted to digitized labor markets and faster job turnover — much less today’s pace of technological change. AI is not only exposing a broken jobs system; it’s revealing a profound weakness in civil society: our inability to work together to solve shared challenges.

The early signs of AI-related job loss are already here. Employers cited AI as the reason for cutting nearly 55,000 jobs in 2025. And already in 2026, Amazon laid off 16,000 corporate employees, in part to free up capital to spend on AI. More troubling are the jobs that never materialized at all: A World Bank study released last November found that job postings in sectors at high risk of AI were down 12 percent since Chat GPT launched in late 2022. And according to the research firm Forrester, AI and automation will eliminate 6.1 percent of all jobs in the United States by 2030. That’s 10.4 million Americans out of work, or 1.7 million more than during the Great Recession.

When jobs disappear, health coverage and economic security often disappear with them, causing widespread damage across families and communities. That’s why good jobs are foundational for a thriving democracy.

Yet according to Gallup, only 40 percent of Americans hold quality jobs, those offering fair pay and benefits, stability, and opportunities to grow. For everyone else, the system breaks down further: training is often disconnected from hiring, employers rely on blunt credentials that screen out capable workers, and benefit cliffs punish upward mobility. Jobseekers are left to navigate a fragmented system largely on their own.

AI didn't create these problems, but it is accelerating them. For decades, economic development and workforce systems were shaped by the decline of manufacturing and the offshoring of jobs — slow-moving disruptions that allowed institutions to adapt over time. AI is different. It is reshaping nearly every industry at once, moving faster than any previous transition, and affecting white- and blue-collar work alike.

Unfortunately, fewer than 10 percent of states have strong economic strategies to translate AI into good jobs, according to BCG Global, and fewer than 20 percent have workforce plans to address its impact. That leaves most Americans vulnerable when (and not if) more significant layoffs begin.

The solution is to build a new social compact — one that treats the ability to move into and between jobs as essential civic infrastructure. While the “abundance” debate focuses on the government’s inability to build physical infrastructure, such as high-speed rail between San Francisco and Los Angeles, we need civic infrastructure nationwide that can move people between jobs just as fast. In an AI-shaped economy, that infrastructure must be designed for continuous transition.

I’ve worked on this challenge from multiple perspectives — architecting the strategy that’s turning Tulsa, Oklahoma into a tech hub, leading a cyber workforce initiative at the New York City Economic Development Corporation, and now supporting The Rockefeller Foundation on its national jobs strategy. Across these experiences, one lesson is consistent: good jobs emerge when civic infrastructure aligns the public, private, and social sectors around work.

Here are four recommendations for building the civic infrastructure a modern jobs system requires:

First, regional economies need to build modern intermediary organizations that can mobilize civil society. Historically, most cities have lacked strong EDOs, and chambers of commerce have rarely met this need. I saw this in northeast Oklahoma, where I founded Tulsa Innovation Labs in 2019, an organization catalyzing up to 70,000 jobs by aligning employers, universities, philanthropy, and government around shared workforce goals. The progress didn’t come from predicting disruption, but from creating a new institution with the authority to coordinate the ecosystem.

Second, economies need shared labor-market intelligence that shows where jobs are growing, which skills employers are actually hiring for, and where displacement is beginning. Recall when I spoke with the mayor worried about AI disruption. The problem wasn’t a lack of concern — it was the absence of a data commons or labor market observatory where employers, educators, and public agencies could pool information and understand workforce change together. Without shared data, institutions work in silos and can’t coordinate effectively.

Third, there needs to be employer-anchored pathways that unite training, credentialing, and hiring into a single process rather than today’s fragmented handoffs. That’s why when I was at the New York City Economic Development Corporation, it was important our workforce training programs were designed in close partnership with industry. When employers help design pathways from the start, reskilling becomes a direct on-ramp to work.

And fourth, our social safety nets must be designed to support mobility rather than prolonged detachment from work. Aligning unemployment insurance, training benefits, childcare, and career navigation around rapid re-employment would turn moments of disruption into fast, supported transitions. In my work with The Rockefeller Foundation, I’ve learned how public policy can help such solutions scale nationally — but only after models are tested across regions and learnings shared with governments that then refine their benefits systems.

Rather than tolerating fragmentation and delay, this approach replaces them with coordination and speed. It treats job transitions as a normal feature of a dynamic economy, not emergencies to be addressed after the fact.

That shift raises an unavoidable question: who pays? As productivity gains increasingly concentrate in a small number of firms building large language models and countless data centers, it’s reasonable to expect those companies to help fund the systems that enable workers to adapt, whether through targeted taxes, data-sharing obligations, or workforce compacts. Absent that investment, the costs of adjustment are borne by workers and communities least able to absorb them.

For the past century, economic development has helped communities navigate economic change. In the decades ahead, its success will depend on whether it can once again reinvent itself — this time to build the civic infrastructure an AI economy requires. Job disruption is inevitable. Building a jobs system that can thrive in the AI era is a choice.

Nicholas Lalla is an urbanist working at the intersection of economic development and emerging technology. He is the author of Reinventing the Heartland (HarperCollins, 2025).