In 1936, Los Angeles County built a bridge across the San Gabriel River to link the San Gabriel Valley with Wrightwood, a popular camping spot. In 1938, a hurricane washed out the unfinished roads on both sides of the bridge.
State and local leaders who build education and training pathways can surely empathize with the engineers and builders of San Gabriel’s bridge to nowhere. Much like the bridge, those pathways may become obsolete in the time it takes to complete them, especially as artificial intelligence accelerates changes to the skills employers demand. National education, training and data leaders are starting to come to grips with the pace and scale of change while developing possible solutions.
It is admittedly unclear how quickly or dramatically A.I. will reshape the labor market. High-profile insiders like the Anthropic founder Dario Amodei predict that A.I. will create rapid job disruptions. Many economists are more circumspect, noting that evidence of such disruptions has been scant so far and that future effects are uncertain. Barriers like skill deficits, weak technical infrastructure, regulatory burdens, or organizational culture can slow A.I.’s adoption in the near term, and opinions diverge widely about its longer-term impact on jobs and skill demands.
Such uncertainty is frustrating for leaders who must plan training systems, workforce support, and career pathways. Some economists argue that the data systems those leaders rely on are struggling to meet the moment. Response rates to federal employment surveys have declined sharply in recent years, undermining confidence in federal data. Long-established methods for projecting labor market trends weren’t built for rapid technological change, and few systems can track how skill demands are changing even if job titles aren’t.
Those challenges create blind spots for state and local leaders who aim to design pathways for opportunity. They need data systems to adapt education and training to new skill demands, identify and help displaced workers, and steer people towards timely and relevant credentials.
There has been some action at the federal level. For example, the White House’s 2025 AI Action Plan proposes that the Department of Labor lead an AI Workforce Research Hub “to evaluate the impact of AI on the labor market and … generate actionable insights to inform workforce and education policy.” Details are pending, but some observers see the hub as an opportunity to make federal data systems more agile.
In addition, states have been collaborating with each other and outside partners to improve labor market information:
- The National Association of State Workforce Agencies (NASWA) received funding to strengthen the National Labor Exchange (NLx), a nonprofit initiative to provide states and employers with accurate and comprehensive real-time data about online job openings. NASWA notes that the funding will help the NLx “respond to the evolving nature of work and the growing adoption of artificial intelligence.”
- Data system leaders in North Carolina, Tennessee and Utah helped validate the Iceberg Index, a metric that models which job tasks AI can perform across 923 occupations. The researchers who developed the index at MIT and the Oak Ridge National Laboratory find that AI’s impact on computing and other technology jobs is the tip of the iceberg. A much larger zone of exposure lies beneath the water’s surface, where A.I. can automate skills in jobs spanning administrative, financial, and professional services. The index aims to help policymakers explore scenarios of A.I.’s impact before committing time and treasure to specific policies.
- Data system leaders in states including Arkansas, Ohio, Oregon and New Jersey are participating in the Industries of Ideas initiative, an effort to understand how public research investments in A.I. affect the workforce in states and communities. Funded by the National Science Foundation, the initiative combines state and private workforce and education data to make sense of the A.I. workforce.
Though still in their early phases, such initiatives are helping state and regional leaders envision data systems that can keep pace with seismic changes.
