The Race — Building Democracy in the Age of AI. People race toward a shared finish line.
The Race
← Trends

Supporting case

Chicago Fed research separates workplace AI use from predictions about jobs

What happened

Kristen Broady’s November 2026 working paper combines historical analysis, the nationally representative 2025 Survey of Household Economics and Decisionmaking, and qualitative evidence from the Federal Reserve’s 2026 Worker Perspectives Project. The publication supplies a month but no verified day; it is recovered research, not a claimed November 7 release.

Among 7,488 workers answering the workplace-use question, the weighted share reporting generative-AI use in the prior month was 24.7%. Among users, 81% agreed that it saved time, while 55.5% said it enabled new tasks. These are reported experiences, not measured time savings. The paper’s tables and limitations ↗ show that its cross-sectional models identify associations, not causes, and do not directly control for occupation, industry, employer policy or task composition. Focus groups deliberately covered workers and job seekers in households earning $85,000 or less; their accounts are not national prevalence estimates.

This gives the implementation-capacity analysis a worker-level counterpart to the OECD firm comparison. It also warns against turning technical capability into a claim that jobs have already disappeared. Workers encountering AI as a useful tool and job seekers encountering automated screening occupy different positions; the evidence does not establish the size of screening discrimination or a causal wage effect.

Dates and implementation stage

Action/evidence 2025 survey / 2026 qualitative evidence; publication 2026-11; approved 2026-11-08

Supporting sources

Related trends