Much of the talk about AI’s potential impact on the labor market has focused on which jobs may be automated or eliminated. But a new analysis zeros in on what some experts increasingly think may be the bigger risk: the disruption of the career pathways that provide economic mobility to millions of workers.

The report, set to be released Thursday by Brookings Metro and Opportunity@Work, warns that AI could not only change individual occupations, but limit income growth for millions of workers. The result, it notes, could “constrain local workforce development, depress economic dynamism, and limit regions’ capacity to adapt to technological change,” with the effects potentially felt nationwide.

“When these pathways break down, both workers and employers are going to feel the bite,” says Mark Muro, a senior fellow at Brookings Metro and the report’s co-author. 

This threat poses a particular challenge to STARs, the 70M U.S. workers without four-year degrees who are Skilled Through Alternative Routes and make up half the nation’s labor force. Their economic mobility depends on transferring their skills from entry-level “origin” roles—like cashiers and receptionists—to higher paying ”gateway” and “destination” jobs. Over the past decade, more than 23M STARs have advanced along such pathways.

But AI may be disrupting that system.

Rather than affecting jobs in isolation, AI technologies likely will impact entire pathways at once. Automating a gateway role, for example, could block advancement not only for workers in that job, but also for those trying to move up into it.

The scale of potential disruption is significant. The report estimates that about 15.6M STARs, roughly one-fifth of that workforce, are in occupations highly exposed to AI. Many of those workers are concentrated in clerical, administrative, and customer service roles that play a central role in career mobility.

It also jeopardizes economic mobility for the 23M STARs with low adaptive capacity, the report notes, since they’re less able to weather job losses and career transitions.

“We need to step back and see what happens when suddenly lower wage workers no longer have that as their next step,” says Justin Heck, senior director of research and data production at Opportunity@Work and the report’s co-author. 

Regional Variations, Employer Impacts

Muro and Heck say geography will play a major role in how these changes unfold, which mirrors other research by Brookings. 

In some metro areas, large shares of workers are concentrated in administrative and service roles with high AI exposure. Other regions may face lower levels of disruption. The report notes that:

  • Career pathways in Northeastern metro areas—such as Albany, N.Y.; Harrisburg, Pa.; and Providence, R.I.—are at greater risk due to the concentration of administrative and clerical jobs vulnerable to automation.
  • Sun Belt-region metros face the highest exposure due to fast-growing service economies that employ STARs in support roles. Florida’s most vulnerable cities include Palm Bay, Cape Coral, Jacksonville, North Port-Sarasota, Orlando, and Tampa.
  • Midwest metros, with their reliance on jobs in operations and logistics, are less exposed, including Cincinnati, Milwaukee, and Des Moines, Iowa.

These variations have piqued the interest of regional intermediaries, who feel “a sense of uncertainty about what are the most important roles to be concerned about and are trying to figure out how to allocate finite resources,” Muro says. 

They also have prompted some regional industry leaders and economic development organizations to think carefully about the impact automation may have on their talent pipelines. Employers rely on these career pathways to develop experienced talent, notes Heck, and disruptions could weaken those pipelines over time. That’s especially true if AI adoption leads primarily to automation without creating new opportunities for skill development and advancement.

“We often don’t acknowledge the fact that your best employee is probably building their skills somewhere else right now,” Heck says. “That mobility between companies creates opportunities for workers, but it also creates future talent.

“Where are we going to find all of the folks with five to 10 years of talent if we’re no longer creating that talent?” 

Finding Solutions

Muro and Heck say that addressing these risks will require policymakers, educators, and employers to shift from a focus on AI’s impact on individual jobs to a broader effort to preserve and strengthen career pathways.

That starts, according to the pair, with understanding where pathways are already beginning to fray. Because AI adoption varies widely by occupation, industry, and region, leaders need better data to identify which routes into higher-wage work are holding up, which are evolving, and which are at risk of disappearing. That kind of visibility could help regions respond before disruptions become entrenched.

Equally important is clarifying which skills are gaining value in an AI-integrated economy and how accessible those skills are to STARs. The report urges moving away from abstract skill frameworks to more practical questions: 

  • What capabilities are employers actually seeking?
  • Can they be learned through work-based experiences or short-term training? 
  • Do existing pathways still allow workers to build and demonstrate those skills over time?

Employers also must develop more intentional approaches to how AI is deployed in the workplace, according to the report. In what they call “high-road” adoption, employers would use AI to help workers learn faster, take on more complex tasks, and move into higher-paying roles rather than simply replacing them. This is the kind of approach that a trio of top labor and tech economists have termed “pro-worker AI.”

For its part, higher education can take more seriously its role providing life-long education, since workers in most fields face a host of technological changes.

The important thing to remember, Heck says, is that “AI’s impact on pathways isn’t pre-set. 

“There are a lot of choices—human choices, organizational choices—that are going to determine how this plays out,” he says.