On-the-Job Exposure to AI Among Lower-Income Workers

Executive Summary

To better understand the potential impacts of generative AI (gen AI) on the economy, this analysis uses quantitative methods to assess the extent to which workers are likely to be exposed to AI on the job, paying particular attention to workers in lower-income households, the occupations and industries in which they work, and how exposure varies across different parts of the country. It also draws on qualitative insights to understand how the impacts of AI integration are showing up in real time and how workforce and training organizations, nonprofits, and employers are adapting.

Key Takeaways

Analysis of 2023 American Community Survey microdata reveals that among workers highly exposed to AI, those in lower-income households:

• Account for more than 6 million (20%) of all AI-exposed workers.
• Are more likely than average to work in Office and Administrative Support occupations and to work in service-oriented industries, such as Health Care and Social Assistance.
• Tend to be older, to have higher levels of educational attainment, and to be higher-earning than lower-income workers as a whole.
• Make up varying shares of the AI-exposed workforce across the country and work in differing mixes of occupations and industries depending on their local labor markets.

Seven roundtable and listening sessions with nearly 60 participants—including employers, workforce system representatives, community-based organizations, and community colleges, among others—yielded the following insights about real-time AI adoption and impacts:

• Adoption and integration of AI into roundtable participants’ organizations and operations varied substantially, but for those integrating AI into their operations, early employment impacts were already apparent.
• Respondents saw many ways in which AI could be beneficial to lower-income workers and job seekers but emphasized the need for critical thinking skills to make AI adoption successful.
• Respondents expressed concerns about uneven impacts of AI adoption worsening outcomes for vulnerable workers, unless adequate guardrails and supports are in place.

Read the full report (pdf, 2 mb)
Download figures and data (xlsx, 57 kb)

The views expressed in this report are those of the authors and do not necessarily reflect the views of the Federal Reserve Bank of San Francisco or the Federal Reserve System.