October 1, 2025
A new report from the Upjohn Institute finds that jobs most exposed to Artificial Intelligence are clustered in white-collar, higher-education roles such as analysts, engineers, and professional service providers. But exposure doesn’t necessarily mean job loss. Many of these occupations are expected to grow as AI augments rather than replaces workers.
In their report, AI Exposure and the Future of Work: Linking Task-Based Measures to U.S. Occupational Employment Projections, Authors Erik Vasilauskas and Mike Horrigan adapt a Pew Research Center task-based framework, using Department of Labor O*NET survey data to measure the importance and intensity of 16 AI-susceptible tasks. They then link those measures with the Bureau of Labor Statistics’ 2023–2033 employment projections, offering a forward-looking view of how AI will influence the job market.
Key findings include:
- High-skill growth: Occupations like accountants, information security analysts, and engineers are projected to expand. In these roles, AI is expected to boost productivity and decision-making.
- Low-skill decline: Clerical and administrative jobs built on routine cognitive tasks—such as data entry, payroll clerks, and customer service representatives—are projected to keep shrinking.
- Mixed outcomes in the middle: Jobs requiring some college or an associate’s degree may grow or decline depending on how much AI complements versus replaces their core tasks.
The report stresses that AI’s impact is not one-size-fits-all. Instead, outcomes depend on education, wages, and whether technology functions as a substitute or a complement to human work. For policymakers and workforce planners, the message is clear: strategies must both support adaptation in high-skill fields and offer transition help for workers in vulnerable roles.
Experts
